Changes in Biogeochemical Cycles

NASA-Earth Observing System

NAG5-6137

Our Interdisciplinary Science investigation addresses the primary biogeochemical cycles of planet Earth and considers, in particular, how they are being changed by humans. The efforts focus on the cycles of water, carbon, nitrogen, and selected trace gases. Process-based models are developed as modules, in concert with database management techniques which synthesize the in situ and remote sensing data needed to characterize regional and global scales.

Models of the Earth's biogeochemical cycles are a central theme. They provide a rigorous means for developing quantitative projections of the interactions of atmospheric composition, climate, terrestrial and aquatic ecosystems, ocean circulation and sea level, and the effects of human activities. The family of models being developed in this investigation provide the predictive link between the physical and biological Earth system and the human dimensions of global change.

The long-term goal of our IDS research is to understand the primary biogeochemical cycles of the planet, the nature of the coupling between the Biogeochemical Subsystem and the Physical-Climate Subsystem, and the characteristics of the human forcing of the Biogeochemical Subsystem and hydrological cycle. Our strategy is to study how element cycles function in natural systems where perturbations in biogeochemical states are driven primarily by climate variability and in systems where disturbance gradients induced by human activity have modified significantly exchanges of water, carbon, nitrogen, or sulfur.

To execute this strategy, we have been developing regional and global, geographically-specific, mathematical models and databases. These describe ecosystem distribution and condition, the biological processes that determine the exchange of CO2 and trace gases with the atmosphere, and the fluxes of carbon and nutrients to aquatic ecosystems. This suite of models rest within interactive information systems that integrate geographic information systems, remote sensing systems, database management systems, graphics packages, and model interface shells. In time a macro Information Management System will be developed to "wrap" the specific subsystem information management systems.

Five science Objectives form the basis for our IDS research.

• To describe the global pattern and distribution of terrestrial ecosystems and to describe the geographical and temporal forcing agents of anthropogenic disturbance, particularly deforestation.
• To describe globally for terrestrial ecosystems key biological processes such as net primary production, heterotrophic respiration, transpiration, nutrient uptake, carbon allocation, and leaching which bear upon global biogeochemical cycles.
• To determine the baseline and changing fluxes of water, carbon, nitrogen and phosphorus from terrestrial biomes to the world's oceans.
• To determine the spatial and temporal patterns of the fluxes in CO2, CH4, CO, N2O, NH3, reduced sulfur gases, and the nonmethane hydrocarbons between terrestrial biomes and the atmosphere, and between aquatic systems and the atmosphere.
• To develop the background trace gas release data associated directly with human industrial and urban activities.

Each Objective is pursued through the development of a set of specific regional-global databases and/or models. This Objective-focused approach is complemented by a set of "cross-cutting" projects which exercise and test various aspects of the models, often at finer scales, with the focus on timely scientific results.

The results of this investigation have supported and will continue to support the priority research needs of the Intergovernmental Panel on Climate Change (IPCC) and other regional and global integrated assessment activities concerning large-scale environmental change.

A. Continental to Global: A Focus on Modeling and Data Mining

a. Terrestrial Ecosystem Model

The Terrestrial Ecosystem Model (TEM) is a process-based model that has been designed to estimate the spatial and temporal distribution of the major carbon and nitrogen fluxes and pool sizes of the terrestrial biosphere at regional and global scales. It was initiated in the first years of our IDS investigation (e.g., Raich et al., 1991), and has continued to evolve through a series of versions. The first two versions of TEM were used to examine spatial patterns of net primary production in South America (Raich et al., 1991) and in North America (McGuire et al., 1992). The third version of TEM was used to examine the response of NPP to elevated temperature and carbon dioxide for temperate forests (McGuire et al., 1993); and to GCM predicted climate change for the terrestrial biosphere (Melillo et al., 1993). The carbon storage predictions of the third version were also evaluated for global terrestrial ecosystems (Melillo et al., 1995) and for grassland and conifer forests (McGuire et al., 1996). In addition, the U.S. Forest Service used NPP estimates from the third version of TEM as part of their 1995 national assessment of the effects of global climate change on forest productivity (Joyce et al., 1995; McGuire and Joyce, 1995; Perez-Garcia et al., in press). Version 4.0 of TEM (McGuire et al., 1995b; McGuire et al., 1997) was developed to improve the dynamics of soil organic carbon along gradients of temperature, moisture and soil texture; and to incorporate the role of nitrogen in the response of forest net primary productivity to elevated atmospheric carbon dioxide (McGuire et al., 1995a). TEM 4.0 estimates of the global soil organic carbon pool (McGuire et al., 1995; Xiao et al., 1997) are less than the estimates of the third version of TEM because TEM 4.0 only considers the pool of fast-cycling or "reactive" soil organic carbon instead of the total amount of carbon stored in the soils.

During 1996, and 1997 we continued to use TEM 4.0 in several regional studies to investigate the potential effects of climate change, elevated CO2 and vegetation redistribution on net primary productivity and carbon storage of high latitude ecosystems (McGuire and Hobbie, 1997), China (see abstract by Pan et al. for the IGBP SAC-IV meeting; Xiao et al., in press) and the United States (Pan, et al., submitted). The results of these equilibrium simulations indicate that terrestrial ecosystems have the potential to act as a net carbon sink if atmospheric CO2 is stabilized and that both ecosystem structure and function play an important role in the ability of terrestrial ecosystems to act as a long-term carbon sink. The TEM 4.0 results also indicate that an interaction between elevated CO2 and climate change may play an important role in the overall response of NPP to climate change. In addition to these regional studies, we also used TEM 4.0 in several studies to examine various sources of uncertainty in developing regional estimates of net primary productivity and carbon storage.

By using TEM 4.0 with various climate and soil texture data sets for the United States, Pan et al. (1996) found that differences in solar radiation data sets had the largest effect on TEM estimates of NPP for the conterminous United States. A similar comparison between VEMAP climate data sets (Kittel et al., 1995), developed for the Vegetation/Ecosystem Modeling and Analysis Project, and climate data sets developed specifically for the northeastern United States (Ollinger et al., 1995) also indicated that differences in solar radiation had the largest effect on TEM estimates of NPP in the northeastern U.S. (see abstract by Jenkins et al. for the 1996 annual meeting of the Ecological Society of America; Jenkins et al., in preparation). Jenkins et al. (submitted) also found that the use of different methods of representing vegetation in a grid cell (i.e., mosaic of vegetation types vs. dominant vegetation) had little effect on regional NPP estimates, but that NPP estimates among grid cells were more variable using the mosaic approach than the dominant vegetation approach.

To develop regional and global estimates of NPP and carbon storage, TEM normally uses geographically-referenced data sets organized at a spatial resolution of 0.5 degree longitude x 0.5 degree latitude to capture the spatial variability of environmental conditions across the region. Because a 0.5 degree grid cell covers a large area and environmental conditions are considered to be constant within the grid cell, NPP estimates might be improved using data sets with a finer spatial resolution. To examine this question, we applied TEM 4.0 to the northeastern U.S. using climate, vegetation, and soil texture data sets organized at a spatial resolution of 2 km x 2 km (see abstract by Jenkins et al. for the 1996 annual meeting of the Ecological Society of America; Jenkins et al., in preparation), and we applied TEM 4.0 to the historical range of temperate forests in North America using data sets organized at a spatial resolution of 10 km x 10 km (see abstract by Nungesser et al. for the 1996 annual meeting of the Ecological Society of America). We found little differences between regional NPP estimates based on the coarser 0.5 degree data sets and the finer resolution data sets. However, large differences in NPP estimates could occur between spatial scales in areas that contained a diversity of vegetation types, such as in mountainous terrain or near ecosystem boundaries. This issue will be revisited with the flight of EOS AM-1 and the resulting MODIS and MISR data streams (see also Section A. d).

We have also used TEM 4.0 in four model comparison projects, two at the global scale and one at the (near) continental scale and one at the regional scale. This last effort begins an active effort to compare modeling efforts and approaches within our IDS team.

Community Terrestrial Biosphere Model Project. In the Community Terrestrial Biosphere Model Project, seasonal variations in net ecosystem production (NEP) estimated by TEM across the globe were used in conjunction with the Max Planck ocean and atmospheric transport models to reproduce the seasonal/latitudinal signature of CO2 in the atmosphere. These results were compared to similar seasonal CO2 reproductions using other terrestrial biosphere models (Heimann et al., in press). For CO2 monitoring stations in the northern hemisphere, TEM 4.0, coupled to the ocean and atmospheric transport models, simulated the amplitude and phase of the seasonal atmospheric CO2 cycle with reasonable accuracy. In the tropics, however, the model tended to over-predict the net seasonal exchanges of carbon. A series of modeling experiments were developed to examine if these differences were due to model shortcomings in the phenology algorithms used, consideration of an inadequate rooting depth in the tropics or the lack of consideration of the effects of land use and vegetation fires on CO2 exchanges when developing NEP estimates. The results of these modeling experiments proved to be inconclusive, but did provide information as to how NDVI data might be used to improve phenology and canopy allocation algorithms in terrestrial biosphere models (Sitch et al., in preparation).

Potsdam '95. The NPP estimates of TEM 4.0 across the globe were also compared to approximately 17 other global models in a model intercomparison workshop (Potsdam '95) sponsored by the International Geosphere-Biosphere Program's (IGBP) Task Force on Global Analysis, Interpretation, and Modeling (GAIM). The global NPP estimates varied by a factor of 2 among the seventeen models (see abstract by Moore et al. for the First GAIM Science Conference; Cramer et al., submitted). Relatively low global NPP was estimated by TEM 4.0 compared to the other global models. Differences in the NPP estimates among the models could be attributed to differences in: 1) the sensitivity of the various models to climate (Ruimy et al., in preparation; Schloss et al., in preparation); 2) calculation of water balance (Churkina et al., in preparation); 3) phenology (Fischer et al., in preparation; Kicklighter et al., submitted) and 4) vegetation distribution used by the modeling groups (Fischer et al., in preparation; Schloss et al., in preparation).

Vegetation/Ecosystem Modeling Analysis Project (VEMAP). For the Vegetation/Ecosystem Modeling Analysis Project (VEMAP), we compared the NPP and carbon storage estimates of the conterminous United States for contemporary climate and three climate change scenarios among TEM 4.0, Century (from the Schimel EOS-IDS Team) and Biome-BGC (from the Running MODIS Team). These biogeochemistry models also used the changes in vegetation distribution predicted by three biogeography models to examine the coupled response of ecosystem structure and function to climate change. Although the biogeochemistry models estimated similar NPP and total carbon storage for the conterminous United States under contemporary conditions, the response of NPP and total carbon storage to climate change varied among the models (VEMAP Members, 1995). All models estimated increases in NPP with climate change and elevated CO2, but TEM predicted the largest increases (up to 40%). The models differed substantially in their estimates of the response of total carbon storage to climate change with TEM predicting increases in carbon storage and Biome-BGC predicting decreases in carbon storage. All three models showed correlations among water use, nitrogen availability and primary production, but the models simulated spatial variability in ecosystem processes in substantially different ways (Schimel et al., 1997). To better understand these different responses, we are currently pursuing a detailed comparison of some of the mechanisms in each of the models such as water balance calculations (Cienciala et al., in preparation; Hibbard et al., in preparation) or how elevated CO2 affects NPP (Pan et al., in press).

TEM, PnET, and DNDC-PnET. In addition to the comparison of global models described above, we are also conducting a formal comparison of TEM to two regional versions of PnET model as well as to a regional hybrid of DNDC and PnET. The original version of PnET was developed specifically for all forest types in the northeastern United States and has been used in our IDS to provide high-resolution estimates of the effect of climate change and atmospheric deposition on the carbon, nitrogen, and water balances of forests in the northeastern United States (Aber et al., 1995). A modified version of PnET has also been developed to provide similar carbon, nitrogen and water balance estimates for forests in the southeastern United States by Dr. Steve McNulty of the U.S. Forest Service. In the northeastern United States, TEM and PnET provide similar regional estimates of NPP, model bias occurs at both the low and high end of the NPP range (see abstract by Jenkins et al. for the 1996 annual meeting of the Ecological Society of America; Jenkins et al., submitted). Although regional estimates of NPP are also similar between TEM and PnET for the entire southeastern United States, TEM estimates a higher NPP for temperate deciduous forests and a lower NPP for warm temperate mixed forests (McNulty et al., in preparation). In addition, NPP estimates for specific 0.5 degree grid cells can differ substantially between the two models in the southeastern United States.

Finally, we are beginning to consider, in this context, a new forest model PnET-DNDC was constructed based on the agricultural version of DNDC, the PnET model. and recent observations at Harvard Forest in the U.S. and Hoglwald Forest in Germany. This model is focused on predicting not only forest photosynthesis, forest growth, C allocation, water and N demand, litter production, decomposition/nitrification/denitrification in the forest floor and mineral soil profile, but also on key N trace gas emissions: i.e., N2O, NO and NH3 from forest ecosystems.

NASA EOS funding has also partially supported the application of TEM as part of an integrated assessment activity being conducted in conjunction with the MIT Joint Program on Global Change. This activity involves coupling TEM with a reduced form GCM, an atmospheric chemistry model and an economics model (Xiao et al., 1995; Xiao et al., 1996a; Prinn et al., 1996; Xiao et al., 1997; Xiao et al., submitted; Prinn et al., submitted). In one analysis of this activity, we found that the linkage of TEM 4.0 to a 2-dimensional climate model was useful for impact assessment and uncertainty analysis within the integrated assessment framework at the scales of the globe, economic regions and biomes (Xiao et al., 1997).

During 1996 and 1997, we have continued to develop a transient version of TEM (Version 4.1) such that stocks and fluxes of carbon and nitrogen in terrestrial ecosystems can fluctuate in response to interannual variability in atmospheric CO2 concentration and climate. The development of TEM 4.1 required the synchronization of water balance calculations with the calculation of carbon and nitrogen fluxes (Melillo et al., in preparation). In earlier, equilibrium versions of TEM, water balance variables (e.g., soil moisture, actual evapotranspiration) were estimated by an intermediate Water Balance Model (WBM, see Section A. c) and these estimates were used as inputs into TEM (see Pan et al., 1996). Version 4.1 is being used to examine interannual variation of net primary production (NPP), net ecosystem production (NEP) and terrestrial carbon storage in response to historical and future changes in atmospheric CO2 concentration and global climate as part of the Carbon Cycle Model Linkage Project (Melillo et al., in preparation; Heimann et al., in press) and the MIT Global Change Joint Program Global System Model (Prinn et al., 1996; Prinn et al., submitted). In addition, a subset of these results has been used to examine interannual variations of NPP, NEP, and carbon storage in the conterminous United States (Tian et al., submitted) and in the Amazon Basin (Tian et al., in preparation). A transient version of TEM will also be used in the upcoming model comparisons of NPP and NEP in Phase II of VEMAP as soon as climate data sets are available.

We have also modified TEM 4.1 to use transient N input data to represent the effect of atmospheric nitrogen deposition. Using the georeferenced N deposition of Dentener (personal communication) and scaling N deposition to historical fossil fuel emission data, we conducted a sensitivity analysis of the effect of ecosystem N retention on the magnitude of annual and seasonal NEP. In historical simulations of 1990, TEM NEP estimates increase from 0.8 Pg C per year to as much as 1.3 Pg for the globe with the addition of atmospheric nitrogen.

TEM has been modified, at the site and transect levels, to simulate the development of grassland, forest or savanna vegetation types, depending on climate, disturbance and soil parameters. We plan to continue this work to include other vegetation types and successional factors so that we can simulate ecosystems as they develop from early stages to mature vegetation and apply these simulations at the regional and global levels.

Other new work will involve additional modifications to TEM 4.1 to improve the interaction of carbon, nitrogen and water fluxes between terrestrial ecosystems and the atmosphere; and between terrestrial and aquatic ecosystems. Proposed model modifications will include changes in how TEM simulates carbon, nitrogen and water dynamics within terrestrial ecosystems; and how TEM is coupled to the atmosphere and aquatic ecosystems. For example, we will modify TEM to explicitly simulate nitrification, denitrification, leaching and biological N fixation. The model will also be able to use atmospheric N deposition prescribed from spatially-explicit data sets or estimated from atmospheric chemistry models. In addition, work will continue on the development of a transient version of TEM that will include changes in both ecosystem structure and function in collaboration with Colin Prentice and his research team from Lund, Sweden.

Currently, TEM does not incorporate land-use and land-use change in its simulations of global carbon and nitrogen dynamics. To assess the effects of human activities on global carbon and nitrogen dynamics, we need to know: 1) where these activities are occurring; and 2) how these activities effect ecosystem dynamics. To determine where agricultural activities are occurring, we are currently developing a cropland distribution model that simulates the spatial distribution of potential croplands under abiotic constraints (see abstract by Tian et al. for the 4th Biennial International Conference for Ecological Economics; Xiao et al., submitted). The distribution of potential croplands will eventually be used as one input into a land use model that defines the spatial distribution of actual cropland. Other inputs to the land use model will include socio-economic constraints (e.g., human population, GNP per capita, crop productivity per unit area, food and nutritional requirements for people). The land cover and land use models will allow us to generate a global land cover data set of natural vegetation, croplands, and urban areas for future and contemporary climate conditions. DNDC will be incorporated into a transient TEM to account for agricultural systems.

We have also started to examine the effects of human activities on carbon dynamics by reviewing recent analyses of the consequences of deforestation on the global carbon budget (Melillo et al., 1996) and examining the consequences of forest-to-pasture conversion on CH4 fluxes in the Brazilian Amazon Basin (Steudler et al., 1996).

b. UVa Terrestrial Model

During 1997, we continued modifications to our primary productivity model to allow more consistent coupling to soil carbon and nitrogen cycling models with respect to daily solution steps. These revisions address a fundamental requirement to treat significant daily variations in primary productivity processes and ecosystem hydrology in conjunction with plant growth, vegetation community dynamics, and changes in soil carbon and nitrogen pools that occur on time scales from decades to centuries. These differences in response times make it difficult to match nitrogen uptake, which varies daily with variations in photosynthetic processes, to nitrogen availability, which depends on the much slower turnover of litter and soil carbon pools.

We completed two analyses meant to clarify spatial and temporal variability in terrestrial primary productivity and the implications of this variability for using satellite remote sensing to observe ecosystem responses to environmental change.

An analysis of spatial patterns focused on the distribution of leaf area across the conterminous 48 U.S. Our model assigns the largest leaf area that satisfies constraints on carbon and hydrologic budgets. As a reference case, we simulated primary productivity and leaf area distributions that satisfy all constraints. We compared this reference case to simulated distributions that satisfy each constraint individually. The VEMAP data set specified climate and soil moisture characteristics. We evaluated sensitivity to atmospheric CO2 increase by generating simulations at pre-industrial (280 ppmv), contemporary (360 ppmv), and elevated (560 ppmv) atmospheric CO2 concentrations. We completed a manuscript reporting results of this study that will be submitted to Global Change Biology in early, 1998 (Emanuel and Woodward 1998).

The simulated pattern of leaf area dictated by the full set of carbon and hydrologic constraints and at pre-industrial or contemporary atmospheric CO2 concentrations corresponds reasonably to natural vegetation patterns and, where contemporary vegetation remains similar to natural vegetation, to patterns in the normalized difference vegetation index derived from AVHRR data. The differences in leaf area distributions at pre-industrial and at contemporary atmospheric CO2 levels are insignificant, suggesting that vegetation responses to CO2 increase thus far are very difficult to detect by satellite remote sensing. However, leaf area is significantly greater at elevated than at contemporary CO2 concentration. The most significant response is in the prairie peninsula region where leaf-area increases from values typical of tallgrass prairie to values corresponding to small stature forests. Generally, increased leaf area at elevated CO2 concentration is due to less limitation by hydrologic constraints associated with increased water use efficiency rather than less restrictive carbon balance constraints.

In a related study, we simulated global leaf area distributions representing climate by each of the two years of data assembled by the ISLSCP project. These simulations contrast leaf area associated with climate typical of El Nino conditions to that associated with more normal conditions. Differences in leaf area between these two sample years were significantly greater than those associated with CO2 increase from pre-industrial to contemporary concentrations. We are now revising a manuscript summarizing these results.

NASA's intensive field campaigns yield unique data for diagnosing models of ecosystem processes. During 1997, Lianhong Gu, a Ph.D. candidate partially supported by this project, applied a distributed canopy process model to analyze vertical gradients in CO2 concentration, temperature, and water vapor pressure observed by the BOREAS field campaign. It is important to determine the influences on photosynthesis of such gradients, which are ignored in ecosystem process models applied at global scales. This analysis incorporated a transfer matrix technique to evaluate long wave radiative transfer within plant canopies. We began to evaluate this technique for use in global-scale simulations. The canopy model and analysis were presented at the 1997 annual meeting of the American Geophysical Union.

An early version of the global primary productivity model developed by this project was included in the Phase I analysis of the VEMAP project, and W. R. Emanuel continues to interact informally with the VEMAP effort. During December, 1996, he arranged and hosted a meeting of the VEMAP group in Charlottesville and contributed to interpretation and discussions of model results from Phase I of the project (Schimel et al. 1997) and to planning for an analysis of model responses to transient climate. W. R. Emanuel also participated in the CMEAL assessment of processes that control ecosystem responses to atmospheric CO2 increase.

We have now completed a reasonably general model of carbon and nitrogen cycling in terrestrial ecosystems that combines detailed descriptions of primary productivity processes, a representation of soil carbon and nitrogen turnover structurally similar to the Century model (Parton et al. 1993), and nitrogen mineralization and limitations on uptake derived from the TEM model (Raich et al. 1991). A suite of analyses is underway to investigate responses to environmental change, and these will continue through 1998. A particular emphasis of these applications is to identify the mechanisms that underlie ecosystem responses that should be detectable by the EOS as data from the AM-1 platform become available.

During 1998, further UVa-model development and testing will concentrate on three areas: (1) plant growth and mortality, (2) seasonal phenology of primary productivity, and (3) natural disturbance, particularly fire. Virtually all terrestrial ecosystem models used to address global change issues require substantial improvement in treating these phenomena. In addition to this work on our basic ecosystem model, we will modify our framework for tracking land use and land cover change in simulations involving ecosystem disturbance by human activities (Emanuel et al. 1994, Emanuel 1996).

Numerous studies demonstrate the importance of plant community dynamics, especially gap phase replacement processes, in determining forest responses to environmental change. However, only models that explicitly treat the establishment, growth, and mortality of individual trees and their interactions satisfactorily represent these community level aspects of forest ecosystem dynamics. These individual based models treat relatively small spatial units, and it remains unclear how dynamics at those scales influence perceived responses to global change such as the EOS will monitor at regional to continental scales. We explored the use of distribution functions to summarize the states of patches on larger land areas. After developing the basic scheme for tracking area in different stages of recovery from disturbance and area undergoing gap replacement, we delayed further development of this approach while completing modules to describe primary productivity processes and soil carbon and nitrogen cycling. We will incorporate our general ecosystem model into the distribution framework during 1998. The initial testing and application of this scheme will be for the eastern U.S. and will seek to clarify how distributions of stand age across the region affect leaf area at spatial scales that are integrated by sensors such as MODIS.

Modifying our primary productivity model to evaluate processes daily substantially improved the realism of simulated large-scale patterns in variables such as leaf area and annual net primary productivity. Daily evaluation is critical to accurate calculation of evapotranspiration. However, data sets that are commonly available for specifying climate in global simulations provide only long-term monthly average values. Thus, we have not fully scrutinized the phenology of primary productivity processes except at a limited number of test sites. Long-term daily records are available for the U.S., and during 1998, we plan to evaluate our model for conditions recorded at a large sample of U.S. stations. We will attempt to compare simulated seasonal phenology with normalized difference vegetation index values derived from AVHRR data and eventually from MODIS data. Initially, we will focus on leaf out and leaf fall in areas dominated by deciduous plants. the mechanisms underlying these critical events are poorly understood and are represented in models only by simple empirical expressions.

Naturally occurring disturbances such as fire can be as important or more so than are climate and edaphic conditions in determining broad-scale patterns in ecosystem characteristics and the distribution of biomes. During 1998, the Uva team will modify their distribution function framework for tracking disturbance by human activities to treat wildfire. Initial applications will be for Southern Africa, again comparing simulations for climate as summarized by the ISLSCP project. We will treat each of the two years of ISLSCP data as if they correspond to long-term average climate, contrasting ecosystem responses to wildfire under climate typical of El Nino events with responses under more normal conditions.

At the leaf level, the dependence of stomatal conductance on photosynthetic rates and on environmental conditions is critical to determining ecosystem responses to environmental change, especially atmospheric CO2 increase. Although there is experimental evidence that leaf water potential and abscisic acid levels affect stomatal responses, the fundamental mechanisms controlling stomates are poorly understood, and most photosynthesis models that are applied globally incorporate empirical representations. Our model uses the function proposed by Ball et al. (1987) to describe the relationship between stomatal conductance, assimilation rate, atmospheric CO2, and relative humidity. Recently, Montieth (1995) argued that including a direct dependence on humidity is inappropriate. Friend (1991) derives stomatal conductance in order to optimize nitrogen allocations. The implications of these alternative stomatal response functions has not been fully investigated.

During 1998, we will assess the implications of stomatal conductance functions for simulated patterns of leaf area and primary productivity. We will simulate leaf area and primary productivity with stomatal conductance described alternatively by the functions proposed by Ball et al. (1987), Montieth (1995), and by Friend (1991). The initial test region will be Southern Africa, and the ISLSCP dataset will specify climate and soil characteristics. In a second phase of the comparison, we will conduct global simulations. However, initially, we plan to constrain model solutions by photosynthesis and conductance data collected in the African test region by Peter Dowty, a Ph.D. candidate supported by NASA's Terrestrial Ecology Program. We plan to submit the results of this analysis for publication by the end of 1998.

W. R. Emanuel continues to collaborate in the development of ecosystem process models with F. Ian Woodward, University of Sheffield. To facilitate this interaction during 1998, Mr. Emanuel will spend one week at the University of Sheffield and Mr. Woodward, a week at the University of Virginia. In keeping with plans to compare stomatal conductance models, most work will focus on improving this aspect of our primary productivity model, which is critical in determining plant responses to atmospheric CO2 increase and under arid conditions.

c. Water Transport and Balance Models.

The Water Balance Model (WBM) forms the basis for the estimation of continental runoff and discharge in our global-scale drainage basin analysis. WBM is also a component of the MBL/UNH Terrestrial Ecosystem Model (TEM) and as such has contributed to our research exploring terrestrial carbon and nitrogen cycling (McGuire etal. 1997, 1993, 1992, Melillo et al. 1993), including the intercomparison of NPP models through the recent Potsdam and VEMAP exercises.

Modeling tests on WBM have centered on the parameter sensitivity of several potential evaporation (Ep) methods performed in a stepwise fashion from point to region to continent and globe. Ep is an important precursor to plant-atmosphere CO2 flux calculations commonly employed in global-scale terrestrial net primary productivity (NPP) models and thus is of direct relevance to our IDS research. Work at the point scale (Federer et al. 1996) centered around a detailed intercomparison of nine methods for calculating evapotranspiration. These methods using daily data were compared for seven sites located from Fairbanks, Alaska to San Juan, Puerto Rico. Three representative cover types were applied at each location. Four of the methods tested were land cover-dependent and these differed by several hundred mm per year for certain location/cover type combinations. A similar finding characterized the cover-independent methods at each location. The normalized seasonal trend in potential evapotranspiration generally corresponded well among the various methods tested. We also tested the use of daily, 5-day, and monthly time steps to drive the PE functions. These findings document both the bias and magnitude of potential errors in water balance components by using time steps greater than one day. Relative to errors in underlying biophysical data fields, this bias is small, on the order of a few percent per year.

These initial studies were carried forward to the scale of the conterminous US (Vorosmarty et al. 1998a). Eleven Ep methods were tested, ranging from simple temperature-driven equations to physically-based combination approaches. Both reference surface (Epr) and land cover-dependent methods (Eps) were evaluated using a priori parameter assignments. The Ep methods were incorporated into the global-scale Water Balance Model (WBM) developed by IDS team members and tested over the conterminous United States (i.e. VEMAP domain) using mean monthly climatic drivers and other biophysical inputs at 30-minute (lat. x long.) spatial scale. For each Ep method water budgets were computed on 3265 individual grid cells using a quasi-daily time step. For 679 locations with reliable discharge measurements and grid-based precipitation we compared simulated to measured annual and monthly runoff. Our objective measure of Ep method performance was mean bias (be) in simulated evapotranspiration (Es) relative to observation, defined as the mean difference between Es and grid-based precipitation minus measured streamflow. Mean bias for individual methods ranged from -94 to +119 mm yr-1 for Epr and -50 to +38 mm yr-1 for Eps methods.

Of what possible consequence are such errors in water balance closure to the accuracy of NPP estimates? In tests using the Terrestrial Ecosystem Model (TEM) across the US, the range in Es bias arising from alternative Ep methods yielded a range in NPP response from 400 to 475 gC m-2 yr-1 (+/- 10 %). This range is nearly identical to that found in recent NPP model intercomparison studies (VEMAP), suggesting that the choice of alternative Ep functions should not significantly amplify inherent uncertainties within such models. The generality of this finding, however, has yet to be tested at the global scale. Our findings also suggest that the calibration and validation of NPP models be conditioned upon comparison of water budget calculations to suitable records of observed discharge.

The Global Hydrological Archive and Analysis System. Our water cycle research relies heavily on a GIS-based software tool developed at the University of New Hampshire. The Global Hydrological Archive and Analysis System (GHAAS) system is cast as a hydrological GIS for multi-scale applications. It has been used in numerous applications within our IDS team including: the intercomparison of evaporation functions for water balance and terrestrial NPP models, determination of US-scale water balances, use of satellite remote sensing to infer tropical river dynamics; and the global impact of reservoir construction on large rivers, all described below. The system has several components including: 1) a meta-database listing both internal and extramural biophysical data sets, 2) a data/model integration package, 3) several input/output functions to manipulate complex spatially-varying input/output fields, 4) resident water balance and fluvial transport models, 5) simulated river networks, and 6) an archive of monitoring data including river discharge (RivDIS v 1.0) and river chemistry (GEMS/GLORI).

Simulated Global River Networks. A major revision of the simulated network topology for world rivers at 30-minute spatial resolution (STN-30 v 4.2) has recently been completed. This data set was prepared through the conjunction of a global digital elevation model (ETOPO5), vector product ARC/WORLD 1:3M stream segments, a routing subroutine embedded within GHAAS. STN-30 depicts approximately 4000 actively-discharging rivers and their associated drainage basins; an additional 1000" potential" drainage basins exist but do not transport water under current climate. We have geographically co-registered STN-30 to a database on more than 1000 monitoring stations (Vorosmarty et al. 1996a) that maintain both discharge and contributing area coverages. Simulated river systems in STN30 v 4.2 show excellent correspondence to reported drainage basin areas gathered from several hundred sites worldwide (UNESCO reporting stations). Linkage of STN-30 to discharge time series facilitates calibration/validation of the Water Balance and Transport Model (WBM/WTM). A recent version of WBM/WTM calculated a global mean for terrestrial runoff estimate within 5% of previous estimates (Vorosmarty et al. 1997a). It is also being exercised in preliminary studies of terrestrially-derived constituent fluxes through our collaboration with M. Meybeck of the University of Paris VI.

Global Impact of River Impoundment on Fluvial Systems. As part of our larger theme of exploring anthropogenic changes to biogeochemical cycles, we are attempting to explicitly consider the widespread alteration of natural river systems arising from water engineering works. Not only do large-scale water engineering schemes have impacts on river chemistry, sediment and carbon flux, but they are also important in defining regional water balances and the character of hydrographs. Corrections must thus be made to account for their impacts when constructing calibration/validation targets for our WBM and WTM simulations.

In a first set of studies (Vorosmarty et al. 1997a, c),the scope and potential impact of reservoir construction on the global network of rivers was analyzed using STN-30 v 4.2, RivDISv 1.0, and WBM/WTM. Storage behind dams represents an eight-fold increase in the standing stock of river water, with residence times for individual impoundments spanning less than one day to several years. The mouths of several of the world's largest rivers show a reservoir-induced aging of runoff in excess of three months. Globally, the mean age of river water has likely tripled to well over one month. The study concluded that the imprint of such storage persists downstream, leading to significant changes in net water balance, flow regime, reoxygenation of surface waters, and sediment transport. The analysis also suggested that the global impact of such changes on suspended sediment and potential carbon flux could be very large. It was estimated that more than 40% of global river discharge is intercepted by large impoundments and that a significant proportion (70%) of this river flow maintains a theoretical trapping efficiency in excess of 50%. The pandemic construction of large dams and other engineering works now represents an important component of the terrestrial water cycle and one that merits appropriate consideration in future global change studies.

In summary, water-related activities in our EOS-IDS investigation have produced models, output, and calibration/validation that help to improve estimation of water pools and fluxes across the continental land mass. This work directly supports our GIS-based modeling of the terrestrial biosphere, most directly through the MBL/UNH Terrestrial Ecosystem Modeling (TEM) effort

d. Modeling Projects with the Schimel IDS-Team

Braswell has taken the lead in our collaboration with the Schimel IDS-Team. Working with B. Holland (and Elaine Matthews, Bill Parton, and J.-F. Lamarque) on a nitrogen deposition project started this year, the team has already made substantial progress analyzing the available spatial/temporal N deposition observations and a plan has been laid out for linking Brasseur et al.'s CTM to CENTURY.

In cooperation with our IDS and the Schimel IDS, Braswell has been developing an operational radiative transfer model inversion algorithm. The goal is to use MODIS and MISR observations together to retrieve canopy parameters (e.g., LAI fAPAR). The algorithm prototype (using NOAA/NASA pathfinder AVHRR data) is currently underway and we expect to have some results 1-2 months before launch. The data ingest routine is done, and there are some other minor details to iron out. Braswell attended the MODIS science team meeting in November and made some very helpful contacts with people, including Eric Vermote, who has offered to provide us with "level 2g" synthetic data to test the algorithm.

Some progress was made in a follow-up to the Science paper (Braswell et al., 1997). Schimel visited UNH earlier this month and Braswell, Linder, and Schimel worked out a simple model framework for doing Kalman filtering and adjoint analysis using the global temperature and CO2 data. The goal is to refine the estimates of the contribution of the terrestrial biosphere to interannual variability of CO2 growth rate, including multiple year time lags. The model works, and we have done some brute-force inversions, so that we have a good start on the more sophisticated statistical analyses.

B. Regional Studies: A Focus on High Spectral Resolution Remote Sensing and Modeling

We have two major efforts addressing the interplay of hyper-spectral observations, environmental issues, modeling: the Aber Team and the Rock Team. These efforts are cooperative and closely related, but they have slightly different thrusts. We shall discuss them separately, but note again that the efforts are highly complementary.

a. Aber Team

Overview. Working with additional support from the multi-agency TECO program, and NASA's Terrestrial Ecology Program, a number of airborne, field, and GIS data products have been acquired to support the objectives of this project. The most important acquisition this year was a spectacularly successful Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) light on 12 August 1997. A total of 84 11x11km scenes were acquired under very rare mid-summer clear sky conditions, including coverage of almost the entire WMNF. Field spectrometer data has been collected at the time of overflight using an ASD FieldSpec instrument. Both sky irradiance and ground target spectra were collected within the study site and will be used as input to atmospheric correction models. This data follows 2 years of AVIRIS data for the intensive study site at the Bartlett Experimental Forest (BEF) within the WMNF. AVIRIS data from the BEF (1995) has been registered to GPS and GIS datasets and atmospheric corrections have been completed. The 1997 dataset for the entire WMNF has just arrived from JPL and processing will begin immediately.

Foliar Chemistry. Foliage sampling has occurred within 2 weeks of each AVIRIS overflight for both the BEF (1995-1997) and WMNF (1996-1997) plots. In addition to the originally scheduled 48 BEF sites and 45 WMNF plots, in 1997 we have been able to incorporate plots from other studies within the WMNF (30 plots established as part of a NASA Fellowship grant (C. Goodale) and 20 plots established by the USFS for a study assessing biogeochemistry at a regional level based on bedrock geology (Scott Bailey)). These additional plots will require only one more visit to obtain the full compliment of data necessary to contribute to the goals of this project.

Foliage samples collected in 1995-1996 (500+/yr) have been analyzed for nitrogen and lignin and cation concentration. Foliage sample numbers have increased in 1997 to 1000+, and are currently being processed for nitrogen and lignin analysis as well as cation content (see collaborations below). The measurement of plot level chemistry has also required field data on foliar biomass distribution by species. This has been accomplished for all sites using either a wide-angle camera point-intercept method, litterfall collection, or both.

We are able to make the distinction of species based on these nitrogen and lignin concentrations. This information will be used to derive estimates of species composition for the extent of the AVIRIS data coverage. Plot level wood production for 47 intensive sampling plots at the Bartlett Experimental Forest has been estimated using whole canopy foliar nitrogen concentration in a simple regression model. These preliminary results suggest that direct measurement of forest canopy chemistry characteristics, based either on field measurements or via remote sensing, may provide simple, direct scalars of current forest productivity potential.

Nitrogen Mineralization. Measurements of net nitrogen mineralization have been made or are currently underway on a total of 18 plots at the Bartlett Experimental Forest between 1996 and 1998. Annual measurements were made on 14 plots for the period of June 1996 through June 1997. For 1997-1998, four new and three replicate plots are currently being sampled. We have used the buried polyethylene bag method with a 6 week in situ incubation period during the growing season plus one overwinter period. For each incubation period, 3 sample pairs are collected from each of 5 subplots at each plot. More than three thousand samples will have been collected by the end of the 1997-98 sampling period.

Samples are processed in the lab and analyzed for extractable nitrate and ammonium. Net mineralization is determined as the difference between the extractable N of the initial and the incubated soil core. A subset of the samples are also being analyzed for ash content, and total carbon, nitrogen and hydrogen content. Foliage from the dominant and co-dominant species of each plot has also been sampled and analyses are underway for nitrogen, lignin and cellulose content.

Data from the 1996 growing season demonstrate that N mineralization was related to both foliar N and the foliar lignin/N ratio. For plots dominated by hardwoods, foliar N alone was the better predictor while the lignin/N ratio was the better predictor for plots dominated by conifers. Plot selection for 1997-98 was aimed at clarifying these trends and improving the distribution of the data.

Historical Land Use Data. Forty-eight maps have been located indicating land use (field, forest, alpine vegetation, or bare) and forest composition (hardwood or spruce) and condition (uncut, partially cut, cutover, second growth, windthrown, or burned) at the time of purchase. The maps were prepared by Federal foresters surveying the land prior to purchase by the federal government (1911-1934) for the creation of the White Mountain National Forest (WMNF). These maps cover nearly all of the present WMNF. Preliminary inspection of the maps suggests that historical fires and uncut forest may cover a greater area than has previously been acknowledged. Several steps have been required to register the historical land use data with existing GIS data planes. On the historical maps, land use was mapped on a tract by tract basis, and these tract boundaries exist today in the U.S. Forest Service's Land Status Atlas. All of the Land Status Atlas tract boundaries have been digitized, edgematched, and registered with existing GIS data planes. Historical tract boundaries and land use data are now being digitized. This information will then be rubbersheeted to fit the tract boundaries as indicated on the Land Status Atlas. Approximately half of the historical maps have been digitized to date, and we expect the process to be complete by early 1998.

Modeling of Forest Response to Environmental Change. In previous years we have published high resolution (1 km) maps on the expected changes in forest productivity and water yield in a double CO2 environment, including effects of climate change and direct effects of increased CO2, as predicted for the northeastern U.S. by the PnET model and nationally and globally from TEM. For a full integrated regional assessment of change, other important stressors need to be added to this analysis. In the past year of we have published the results of two additions to the PnET models: a complete N cycling algorithm, and an ozone effect routine. These additions allow us to address the effects of excess N availability (N saturation) and tropospheric ozone.

N saturation. PnET-CN adds a complete nitrogen cycle to the initial C and water balances in PnET-II, thus allowing the analysis of the effects of N deposition and saturation in this region on C and water balances. In a first paper (Aber et al. 1997) the N cycling module was presented and the full model was validates against measured patterns in mean monthly, and interannual total, nitrate leaching from the control watershed (W6) at Hubbard Brook, NH. We then examine the role of land use in determining decadal patterns of nitrate loss, and conclude that major pulse events such as fires, hurricanes, conversion to agriculture and forest harvesting can effect N cycling for 2 centuries or more (figure 4). Maximum sustainable rates of N cycling for forest ecosystems within the northeastern US region were mapped by developing relationships between this value and summary climatic variables.

In a second paper (Aber and Driscoll 1997), an algorithm was added describing the effect of soil moisture content on N mineralization rates. This improved the fit between model predictions and measurements of total annual nitrate leaching (Figure 5). The model was then tested against transient N leaching data for several disturbed watersheds at Hubbard Brook and surrounding sites. In general, the model predicts temporal patterns in response to disturbance, and clearly predicts and separates patterns in old-growth forests from those of younger stands. The model is then used to predict the effective change in total carbon storage within these forests resulting from increased N deposition. The C:N ratio of increased storage is similar to those predicted by a global perturbation model (Townsend et al. 1997), but the total C storage is lower because the mean retention of total N added was closer to 50% than the 80% value used by Townsend et al. (figure 6).

Tropospheric ozone. Ozone effects were added to PnET-II (Ollinger et al. 1997) through the development of a set of algorithms which describe the attenuation of zone concentrations within forest canopies, based on tower measurements within a deciduous forest canopy at the Harvard Forest, and the cumulative effects of ozone exposure on the photosynthetic potential of foliage. This latter scalar is related to canopy conductance such that faster growing forests realize greater growth reductions, and that drought episodes tend to diminish ozone damage. Model estimates for the northeastern US suggest that ambient levels of ozone resulting from current air pollution conditions reduce forest wood production by approximately 10% per year. This varies year to year and also from site to site. We are currently investigating various hypotheses regarding interactions of ozone effects with carbon allocation and stomatal control. This will be reported on in our 1998 Report.

b. Rock Team

Overview. The Rock IDS Team's research activities continue to focus on developing methods of improving our ability to detect, map and monitor forest damage caused, at least in part, by air pollution (ozone, SO2, and cloud and precipitation chemistry). In particular, the use of narrow-band spectral (hyper-spectral) data to detect the early stages of physiological damage, as well as higher spatial resolution data for the purpose of unmixing the multiple spectral contributions of various scene components to hyper-spectral pixels, were of primary interest. Principal study sites are in the Czech Republic (the Krusne hory and Sumava Mountains) and New Hampshire (the White Mountain National Forest). The outreach education activities focused on incorporating K-12th grade teachers and students, on both a local and an international scale, into research activities of the EOS team members (both at UNH and the larger EOS research community). In addition, members of the EOS IDS team were active participants in an outreach effort designed to introduce private and public sectors to likely climate change impacts in the New England region (including upstate New York). Presentations of EOS IDS research activities and modeled products (PnET through the efforts of the Aber Team) were made as part of the New England Climate Change Impacts Workshop, held at UNH, 3-5 September 1997.

Czech Republic Research Activities. Previous studies involving the use of Landsat Thematic Mapper (TM) data to assess and characterize levels of forest damage in the Czech Republic and the Federal Republic of Germany have demonstrated that only three levels of forest damage (light, moderate, and heavy damage) can be discriminated using regression methods. Based on ground assessment methods employed by Czech and German foresters, a minimum of five levels of damage can be recognized in the Krusne hory/Erzgebirge region of the Black Triangle in central Europe, northwest of Prague. Because of the spectral similarities among the lightest levels of damage, the current TM does not provide adequate discrimination capabilities needed for detection of the earliest stages of forest damage. In addition to study sites representative of a range of damage conditions, the most healthy forest stands must also be identified and characterized. For this purpose, the Sumava Mountains, to the southwest of Prague, was selected to represent very healthy Norway spruce stands.

In addition, as the result of continuing assessment of forest conditions in the Czech Republic (supported in part, by EOS IDS funding), a comparison of field spectral measurements made of Norway spruce in the Krusne hory in 1991 and 1995, strong evidence of forest recovery has be discovered. Field visits to the Krusne hory, in November, 1996 and again in August, 1997, have confirmed the dramatic recovery of previously heavily damaged Norway spruce stands, apparently in response to political changes, and subsequent changes in environmental efforts in the Czech Republic, following the 1989 "Velvet Revolution." Since detection and identification of the initial stages of forest damage, as well as initial stages of recovery, are essential for monitoring forest conditions and treating forest damage conditions effectively, the EOS IDS 1997 activities have focused on:

• Identifying and characterizing a large number of study sites (166) that are typical of a large range of damage conditions in Norway spruce stands, from healthy to heavily damaged;
• Identifying the spectral fine-features characterizing the initial stages of damage in Norway spruce, as well as recovering trees, in both heavily damage stands and healthy stands;
• Identifying the spectral fine-features characterizing the background conditions (understory components) associated with initial damage, as well as recovering trees;
• Using narrow-band (10nm-wide interference filters) video images of the spectral region known as the red edge (shown to be the spectral fine-feature characterizing the initial stages of damage); and
• In coordination with researchers at Charles University, in Prague, detailed measurements of ground parameters (needle chlorophyll content, canopy condition, spectral characteristics, etc.) will be conducted of all 166 field sites.

In the summer of 1997, a UNH research team conduced a detailed field study in the Krusne hory, and the Sumava Mountains (heavily damaged and remarkably healthy Norway spruce stands in the Czech Republic), to utilize a standard methodology for field evaluation and classification of stand-level damage. The relationships between the spectral characteristics of forest stands (using TM data and a field spectroradiometer) and the auxiliary ground data and damage estimates developed in 1995 (as part of Rock's EOS IDS activities) were implemented. Current information on the magnitude and distribution of forest damage, as well as all of the other field measurements, in the two mountainous regions were obtained at a total of 18 sites (15 sites in the Krusne hory, 3 in the Sumava Mountains). Additional sites have since been established in each of these areas by our Czech colleagues from Charles University.

Based on large scale aerial color-infrared photography (1:5000, CIR KODAK 2443) and the German photointerpretation manual classification of the damage characteristics was performed to determine the stage of decline of relatively homogeneous large tracks (larger than 5 ha.) of spruce forest. Using this method 15 stands with comparable elevation, climate and soil conditions, were selected as representing the five different damage classes (3 stands each), as well as the differences in spectral and texture properties. Field assessment was performed at each site noting the damage class, and representative areas - homogeneous and covering the typical variation within the stand, were identified. Within a sampling plot, we characterized stand structure, estimated the damage class of selected trees, and recorded auxiliary data by following standard field procedures and measurements. Spectral reflectance data for the dominant scene components (overstory and understory) were acquired using a GER VIRIS (Visible/Infrared Intelligent Spectrometer) and the Stennis Space Center's Narrow-Band Camera (NBC) system, developed by Dr. Greg Carter (NASA/SSC).

All plots were precisely located, using forest survey maps and a GPS unit. Standard forestry measurements were made within a one-tenth acre plot. To determine the damage class of each site, the German standardized methodology was followed and data were recorded. A total of three study sites, all representing very healthy conditions, were established in the Sumava Mountains.

At one field site (Horni Blatna), in the western Krusne hory (a relatively healthy region), a communications tower was used to acquire Stennis NBC images of pixel-sized areas from an approximate height of 50m above canopy. Two damage conditions, "health" and "early damage," were imaged and "pixel" components converted to reflectance values using ground calibration targets. Similar tower measurement assessments were attempted at other study sites, but could not be used for pixel modeling, due to local geometry between towers and forest stands (too oblique).

Preliminary findings from the 1997 field study suggest that:

• Canopy closure decreases with an increase in damage class;
• Variations within canopy density generally increased with an increase in damage class (PAR data);
• Overall the percentage of ground cover increases with an increase in damage class due to a decrease in canopy closure, but there are high variations due to the particular micro-environmental conditions;
• The differences in height among stands from different damage classes are not significant; and
• Stands with similar elevations (950-1060m) slopes, and aspects exhibit different stages of decline, so elevation is not the main driving force of the decline and site micro-environmental conditions must be more fully explored.

Clear spectral and cellular-level evidence is present for documenting the initial stages of forest recovery, confirming earlier observations resulting from EOS IDS work. Anatomical assessment and chlorophyll concentration analysis activities continue at the present time, and the collaboration with Dr. Jana Albrechtova, plant physiologist and plant anatomist at Charles University, Prague, and her doctoral student, Ms. Jitka Bilkova, is most active. Ms. Petya Entcheva, a doctoral student at UNH, is actively coordinating plans for the upcoming field season in the Czech Republic.

Current data analyses are focused on:

• Spectral characterizations of both VIRIS reflectance measurements and the Stennis NBC imaged brightness values converted to reflectance;
• Microtechnical processing and anatomical assessment of 1st-, 2nd-, and 3rd-year needle thin sections for Norway spruce;
• Chlorophyll determinations for the same needle samples used for spectrometry and anatomical assessments;
• Digital image capture, digital processing and image analysis of selected needle cross sections; and
• Modeling of the pixel mixing/spectral contributions of sub-pixel components.

Ms. Jitka Bilkova (doctoral student at Charles University, Prague), Ms. Petya Entcheva (UNH doctoral student) and B. Rock presented preliminary results of 1996 field, spectral and anatomical assessments at an international research symposium ("Holistic Assessment of Landscapes"), held in Moldava, Czech Republic, in May, 1997. As a result of these presentations, Dr. Tomas Burian, Director of the Czech Department of Environmental Protection in Teplice, agreed to assist in the planning and coordination of the 1998 airborne campaign in the Czech Republic.

Due to the loss of the Lewis hyper-spectral mission, no space-borne hyper-spectral data sets will be acquired for the Czech sites in summer, 1998. However, arrangements have been made to acquire at least two types of airborne hyper-spectral data for these sites. Czech aircraft will be used to acquire NBC-style video coverage for selected field sites, using the Stennis RDACS-1 3 camera system, similar to the NBC field system used in 1997. The same aircraft will be used to fly the Goddard ASAS hyper-spectral platform for the same sites. The Czech Forest Inventory Program (LesProjekt) will also fly the same sites using large scale aerial color-infrared photography (1:5000, CIR KODAK 2443), as part of this joint airborne campaign in the Czech Republic.

Restoration of the north Bohemian black triangle. While participating in the "Holistic Assessment of Landscapes" Symposium, mentioned above, a new research opportunity was discussed. One goal of the Czech government is the restoration of parts of the north Bohemian black triangle area - of both the open-cast mining areas, situated near Teplice, Most, and Chomutov, and the large-scale forest dieback, in the Krusne Hory mountains north of this brown coal mining basin. The landscape in this region has been severely damaged - 30,000 ha of forest had to be cut because of severely damaged trees since the 1960s, with 52% of the forests lost on the Czech side of the Krusne Hory mountains (as determined via past EOS IDS research using Landsat TM data). The resulting clearcut areas have twice undergone unsuccessful attempts at reforestation. Exposed open cast-mining areas total 3600 ha (the Most basin covers 1450 km2 with 12% of the basin presently considered as directly impacted by open cast mining activities). In this area, over 19,000 people were displaced as environmental refugees as the result of the massive strip mining activities.

The problem to be addressed by an international team of research scientists (including B. Rock, and others attending the Symposium) is how best to restore the functionality of the landscape - in order to reestablish both the water cycle and the hydrologic carrying capacity of this area, as well as the forested areas. It will be necessary to follow a holistic approach, assessing the hydrologic regime and the forest conditions at the same time. We do not consider the problems of forest dieback and open-cast mining as separate issues, but rather interconnected issues (in terms of both being parts of an interrupted water cycle) to be addressed together as different aspects of the related issue of sustainable management of the landscape. Such an integrated landscape management effort must address the issue of restoring the functionality of landscape - which includes re-establishing a functioning water cycle.

The restoration of the water cycle for this area is vitally important. A functional, actively transpiring vegetation cover is essential for adequate temperature stabilization of the landscape. Reforestation attempts will not succeed unless microclimatic conditions are locally stabilized and restored by means of re-establishing the functionality of the water cycle. Open cast mines, some proposed to be up to 200m in depth, has resulted in a dramatic suppression of the regional water table. The emissions of high levels of SO2 and heavy wet/dry deposition from coal-fired power plants co-located at the open cast mines have impacted both forest conditions and local ground water quality. In addition many natural lakes and wetland areas have disappeared, or have been heavily modified as have local rivers and streams. The resulting loss of forests, combined with lack of available groundwater near the surface, has reduced total evapotranspiration over extensive areas and hence the landscape's cooling capacity. Large temperature amplitudes which correlate well with landscape patterns can be clearly seen using optical and thermal data acquired from Earth-orbiting satellites.

Several mines will be closing in the near future (1998/98), depending on the economic situation and political decisions. One reclamation approach is to reclaim the closed pits as land suitable for agriculture by infilling these large holes with overburden and transported fill. Huge amounts of material (including mine tailings) would need to be transported to the empty mines, and the acidification of ground water formed by oxidation of sulfur compounds in these tailings would result in serious soil damage. A second approach involves filling the mines with water, and assuming a sensible strategy is adopted, would avoid problems of soil acidification, the considerable costs of material transport, and most importantly would assist in re-establishing the water regime of the whole area. In the Most Basin, five out of six open-cast pits are proposed to be filled with water. The final water level, bottom morphology, and strategy of filling the individual mines should be clearly identified in advance, in order to prevent erosion, acidification of water, and to achieve the best and most sustainable water quality. The soil ground water table will rise again, decreasing the rate of soil organic matter mineralization, and result in decreased transport of matter from the old mining area into surface waters.

The international team proposes to use a multidisciplinary approach based on access to EOS-AM platform sensors (ASTER and MODIS) to monitor and characterize optical, thermal and microwave surface properties associated with the temporal and spatial distribution of water cycle processes in the investigation area. Using this holistic view, the status of vegetation, and the connected functions of the water cycle (evaporation, condensation and precipitation) together with temperature and the soil processes of stabilization and leaching, can be overviewed. Spatially and temporally-related measures can be formulated from high resolution analysis of the landscape, and a detailed feedback of the success of measures undertaken can be achieved on a functional basis.

The following point measurements will be conducted, mostly with high time resolution:

• daily sum of precipitation;
• temperature at different levels in the soil (10 cm.), at soil surface, and 10 cm. and 100 cm. above in air, with a time resolution of 20 minutes;
• level of water in soil ("ground water table") at 20 minutes intervals;
• run-off and electrical conductivity of water transport systems (streams and rivers in selected catchments), at 20 minutes intervals, as measure of base cation losses from the landscape;
• water chemistry analysis in the same streams and rivers, at about monthly intervals; and
• history of the landscape can be investigated with sediment cores from bogs or lakes (pollen analysis for type of vegetation and dating; chemical analysis reflecting soil, water and bog conditions).

The sampling sites will be situated in a transect from the floor of the Most basin up to the top plateau of the Krusne Hory mountains. Reference sites with healthy forest, possibly including the German part of Krusne Hory, will be included. Landscape level estimates will be conducted, including landscape cover type and spatial analysis, vegetation health status, and thermal changes at a landscape level. For the refilling of pit excavations, little water is available from local rivers. Therefore the tailing piles, and also in part the slopes of the pits, should be covered by vegetation as soon as possible. The maximum amount of evaporating vegetation will help to stabilize the water cycle in the area, particularly as the ground water level has been decreased to such a great extent.

The quality of water to be used for filling, in respect of eutrophicating nutrients during the phase of filling, is less critical than the inflow of nutrient-rich water after the pits are already filled. The management of the water quality during the filling, and after filling, has to take into account lake morphometry, the forming littoral zone, the possible water quality and quantity of the inflowing water, and effective in-lake measures. A wide spectrum of different water qualities and uses is possible through early and careful planning of lake filling and management.

The team proposes to use a variety of remote sensing approaches, including the use of high spatial resolution optical (30m pixels) Landsat Thematic Mapper (TM) data, thermal data (from both ASTER and MODIS), and passive microwave data from Special Sensor/Microwave Imager (SSMI). We will use these sensors to monitor transpiration (canopy moisture and resulting atmospheric water vapor) and the resulting cooling of the canopy, as well as the thermal footprint associated with mining reclamation efforts.

In addition to satellite-based measurements and the extensive ground-based survey measurements, there are additional ground-based measurements that will also be required at selected points within each catchment, including the following:

• Maximum and minimum air temperature measured daily, at 1.35m above the ground;
• precipitation (both solid and liquid);
• soil moisture at 15, 30, 60, and 90 cm. depth, on a weekly basis;
• vegetation of the surface (species type, frequency of occurrence, density
• of canopy and ground cover);
• IPAR (Incident Photosynthetically Active Radiation);
• litter chemistry; and
• past land-use history.

These ground measurements, combined with the spatial-distribution of dominant land cover types mapped using the Landsat TM data, will serve as input to the PnET model as a means of modeling the photosynthetic and transpiration capabilities of the landscape. In addition, the project shall use the modeling and data base approaches developed by Charles Vorosmarty, our IDS college. Many of these measurements (air temperatures, precipitation, soil moisture, vegetation condition, and past land-use history) could be collected by students from GLOBE schools located within the catchment basins. Where available, basic meteorological data from Czech sources will be used as well.

Measurement of discharge of streams in relevant catchment and sub-catchment areas in order to show water flow rates and dissolved mineral ion losses. Automatic measuring sondes can be installed at relatively little cost to measure flow rate (from pressure sondes monitoring water levels corresponding to calibrated flow rates) and total dissolved mineral salt losses (by means of conductivity sondes calibrated by periodic laboratory analysis of water chemistry). This is required in order to estimate landscape efficiency in terms of the degree of closed water and matter cycles, compared to open cycle chemical-reaction processes which result in irreversible losses from the landscape and thus lower landscape (ecosystem) efficiency.

High resolution temperature measuring sondes placed in strategic habitats can show temporal patterns of temperature variation and can be used to relate to spatial thermal data achieved from remote-sensing data. Together a thermal landscape efficiency can be calculated for different areas and be compared to the chemical landscape efficiency estimates.

C. EOS AM-1: Instrument-Focused Activities

The MODIS (Moderate Resolution Imaging Spectroradiometer) Land Cover/Land Cover Change product to be produced quarterly at a global scale (1 km. resolution) is a crucial component of EOS and global change research. It is a required input to other MODIS Land (MODLAND) modeled products such as BRDF (bi-directional reflectance distribution function), VI (vegetation index) and LAI (leaf area index).

We are focused on several aspects of MODIS and other sister instruments on EOS AM-1. Amongst those issues are : MODIS-MISR for NPP (See Section A. d) and subsampling on higher spatial sensors. This latter topic and the varies dimensions of the problem are discussed here. The initial focus is on training, validation and accuracy assessment of global land cover.

Unfortunately, there are not, however, adequate requirements and specifications currently developed for training, validation and accuracy assessment of global land cover. The primary restriction to testing and validating global land cover products (maps) is the lack of an adequate and accurate land cover database for training and testing of a site data acquisition and archive system. At 1 km., there are over 150 million land pixels for each global acquisition at this resolution. A one-percent sample would require 1,500,000 observations for accuracy assessment.

MODLAND research teams have been developing an integrated validation strategy based on a hierarchy of global sites selected on the basis of instrumentation and capability level of observation. These comprise a small number (10-20) fully-instrumented tower sites, a larger number (20) of partially-instrumented sites, 8-10 field campaign sites such as BOREAS (BOReal Ecosystem-Atmosphere Study), LBA and SALSA field sites (50) and finally, a large number of sites having remote sensing capabilities (projected at some 400-600), only some of which may be validated by field observation. The basal tier of global remote sensing sites is the most difficult to implement at present due to the cost of supporting data acquisition and feature extraction. These sites must represent global climatic, geographic and biological variability.

A paramount issue in the development of a functional training, testing and validation system for global land cover is the creation of an operational database. While a multitude of global sites exist, information for these sites has not been validated or extracted in any sort of standard format, therefore the cost-added to produce the needed informational and format continuity is substantial.

Training and Validation Issues. Test sites are promoted to take advantage of cost savings by using remote sensing data that have already been generated and where research is continuing. This approach is not however fully developed; the test site data are not in any standard format (classification system or data collection or archiving), they are derived from a number of remote sensing systems and other sources and may be invalidated. However, if the test sites are reasonably representative of their region (an unknown), test site statistics can at least indicate weaknesses and strengths in the datasets and allow users to anticipate how errors might impact their own research. The use of invalidated, biased test site data for accuracy assessment is an extremely subjective and qualitative approach. It should not be considered in the same category as development of an objective, statistically valid sampling system.

Reference land cover data for training and validation are generally pre-existing maps, ancillary biogeophysical data, remote sensing data and test site field data. The quality and availability of adequate training/validation data derived from field sites and existing maps and tabular data is crucial to landcover and landcover change validation. This quality of reference map data is a function of their inherent locational and thematic accuracy. For individual test sites, the utility and quality of the ground truth are variable. Data utility is influenced by the classification system that is applied to the input parameters that are derived for a site, the data source (satellite-derived, field studies, etc.) as well as the level of training of the observer.

The factor which primarily affects the quality of reference data is the underlying accuracy of the ground truth classification which may not be known. The time differences between the collection of source data used to develop a reference dataset, and the remote sensing acquisition date(s) impact both utility and validity of test site data. Because factors relevant to validation vary considerably from test site to test site, validation requires assessment of each test site and most probably re-analyzing the site data to extract specific landcover and landcover change parameters. This argues strongly for the development of a high spatial resolution reference dataset that is derived from other remote sensing sources such as Landsat Thematic Mapper (30m pixels) or SPOT (10 or 20m pixels), standardization of the classification system used, standardization of procedures (protocols) for deriving classification subunits and parameters, and development of a global sampling scheme and associated database.

High spatial resolution imagery will be available from a number of sources including ASTER (which will be on the EOS-AM platform with MODIS) and the Landsat-7 ETM instrument (due for launch at about the time of the EOS-AM platform and destined for a near-simultaneous orbit). Using these fine-scale data, spatial heterogeneity within the test sites can be further characterized and monitored. As a continuing data source, these instruments will also allow updating of landcover ground truth at test sites through the EOS era (2000-2020). Use of collateral remote sensing datasets such as TM provide the additional benefits of redundancy and complementarity that are required by data integration and data fusion techniques.

Thus, the test sites are not available globally, even if they were developed to be useful for comparison with MODIS landcover at a local scale. A substantial effort is required to derive consistent units for these sites to become valid input a global site database system and as a tier in a multistage sampling scheme.

D. Graduate Education

Our most recent IDS-supported Ph.D. students completed their studies in 1997. Rob Braswell has joined the IDS team as a post-doctoral fellow. His scientific contributions are discussed else where in this Report. The other student, Keshav Sharma, currently employed in the hydrometeorolgical service of Nepal, performed detailed studies of the Kosi Basin of the Himalayan region to identify the relative impact of climate vs. land cover change on water balance and suspended sediment flux (Sharma 1997). His study also employed the WBM, in conjunction with time series analysis and statistical correlation techniques. The study demonstrated that the predominant influence on distributed water balance and the nature of observed hydrographs across the region was mainly due to within-basin land surface characteristics and the spatial pattern of climate drivers, not progressive climate change as depicted in the observational record over the last 30 to 50 years. Potential climate-induced increases in snow/ice melt, however, rivals projected land cover changes as the most important factor regulating future sediment delivery.

Ms. Jitka Bilkova (doctoral student at Charles University, Prague; see again Section B. b), is in residence at UNH during the 1997/97 academic year. While at UNH, she will take course-work in aerial photo assessment and photogrammetry, digital image processing, GIS, electron microscopy and analytical field methods. These are courses, unavailable at Charles University, that complement her doctoral study program.

E. Public Out-Reach

The impact of EOS IDS faculty and research activities extend far beyond the UNH campus.

a. Meeting the K-12 Educational Challenge.

In addition to the graduate and undergraduate courses offered through the University, B. Rock has developed several environmental outreach programs designed to introduce both K-12 teachers and their students to field, laboratory, and satellite data analysis methods for assessing their local environment. A New England-wide program, entitled Forest Watch, provides training workshops which are designed to assist elementary, middle and high school teachers in introducing their students to selected hands-on techniques, based on University research methods, for evaluating the health of white pine (Pinus strobus), a known bio-indicator for tropospheric ozone damage.

In the process of learning specific measurement and analysis techniques, students are also exposed to global environmental issues, high-tech methods and equipment for assessing environmental conditions, and practical ways of solving many environmental problems impacting their lives. Patterned after Forest Watch, the GLOBE Program is an international environmental science and education program which creates a partnership between students, their teachers, and the scientific research community. The students participating in GLOBE make measurements of selected atmospheric, hydrologic, and biologic parameters, following protocols developed by the research community, which in turn, uses these student-collected datasets in their own research. EOS and UNH faculty continue to have a major impact of the development of this Vice Presidential science education initiative.

Forest Watch. During 1997, the Forest Watch Program was extended into Vermont, Massachusetts, and Connecticut, as well as expanded within New Hampshire and Maine. A specific educational module, focused on advance remote sensing platforms such as the EOS-AM Platform, was developed to introduce teachers and their students to two new sensors which will be used for improved land cover mapping and environmental assessment activities. Both MODIS and ASTER sensor systems are described, both in terms of their relative capabilities and their likely products. The Forest Watch teachers are anxiously awaiting the scheduled mid-1998 launch of the EOS-AM platform.

A similar educational module, centered on the impending launch of the Lewis spacecraft was also developed for Forest Watch teachers. The write-up compared the partnership developed between Captains Meriwether Lewis (the scientist) and William Clark (the expert woodsman and "pathfinder") as an example of the importance of coordinating scientific investigations with detailed knowledge of ground conditions. Both men truly needed the other, and the expedition would not have succeeded if both had not participated. It was a true partnership, in much the same way that both Forest Watch and GLOBE are true partnerships between the research interests of EOS scientists and the knowledge of ground conditions provided by local schools. The module unfortunately made reference to the Lewis HSI (Hyper-spectral Imager) and the Clark hyper-spatial sensors. With the failure of Lewis spacecraft to achieve orbit, the module has been put on hold. The MultiSpec software that is used by teachers and students in Forest Watch and GLOBE will eventually be used with Lewis and Clark datasets. With the successful launches of both craft in the future, students and teachers will become part of the new Lewis and Clark expedition!

An effort continues to adapt the PnET model to a form useable in the middle school and high school classroom. This activity is being coordinated with J. Aber (UNH EOS IDS Team member).

GLOBE/MODIS. The University of New Hampshire (UNH) has developed a collaboration between NASA's MODIS Land Cover Modeling team, Boston University and the GLOBE Program to make use of GLOBE land cover maps. Validated GLOBE land cover maps, generated by students for their 15 X 15 km. GLOBE Study Sites, based on Landsat TM data and using the Modified UNESCO Classification (MUC) system, will represent such a standardized and validated land cover dataset. UNH will use the existing GLOBE Student Data Server to provide the MODIS Modeling Team with timely access to the student-generated land cover data. In turn, the MODIS Modeling Team will use GLOBE land cover data from selected ecoregions for both calibration and validation of MODIS land cover products. UNH will develop supplemental education modules, including learning activities, for use by GLOBE schools, to explain the significance of their land cover data, how the data will be used by the MODIS Modeling Team, and what the final MODIS land cover products will look like. In this way, GLOBE student data will provide the international research community with unprecedented access to validated land cover data, acquired within the 90m X 90m GLOBE land cover sample sites, and classified according to a single classification system (MUC).

Although a high degree of confidence and a quantified statistical accuracy will be associated with the GLOBE data, Boston University has been developing techniques for evaluating the integrity of training data. Using complementary multistage sampling, we believe that the GLOBE data can provide a standardized and reliable source of land cover data for product validation.

b. New England Climate Change Impacts Workshop - The New England Regional.

Climate Change Impacts Workshop, hosted by the University of New Hampshire's Institute for the Study of Earth, Oceans, and Space, was held at the New England Center on the UNH campus in Durham, New Hampshire from September 3-5, 1997. This workshop, one of 17 regional workshops to be held around the U.S. over the next two years, was seen as an excellent opportunity to inform the general public and state and local agencies about how remote sensing, modeling and associated technologies can be used as effective environmental monitoring tools. Members of the UNH EOS IDS (B. Moore, J. Melillo, J. Aber, B. Rock) team served as key participants of both the workshop organizing committee and the scientific panels assembled to present up-to-date findings regarding the impacts climate change may have on the New England region.

A total of 122 participants, representing a broad range of stakeholders from all the New England states plus upstate New York, attended the first two days of the workshop. September 5th was a writing day involving breakout session leaders, reporters and facilitators focused on production of a draft version of the Summary Report. Representatives from each of the seven sectoral breakout groups (Natural Resources, Human Health, Business and Insurance, Energy and Utilities, Government and Resource Management, Recreation and Tourism, and Information Transfer) have reviewed and contributed to the final version of the report. The New England Regional Workshop Summary Report will provide input to the National Assessment Report.

F. Data Archive Contributions and DAAC interaction

As we begin to get closer to launch of AM-1, we have increased our efforst to work closely with the DAACs and other data issues.

a. The Oak Ridge National Laboratory DAAC

W. R. Emanuel chairs the User Working Group of NASA's Biogeochemical Dynamics DAAC at the Oak Ridge National Laboratory. Vorosmarty also serves as a member of the Oak Ridge DAAC User Working Group.

Global River Flux Inventory. A first version of the Global River Discharge Database (RivDISv1.0) was recently published as part of the UNESCO Technical Documents in Hydrology series (Vorosmarty et al. 1996a). Time series data are presented on nearly 1000 rivers worldwide, providing us with important calibration and validation targets for the WBM and DBM models. The data are part of the larger Global Hydrological Archive and Analysis System (GHAAS) developed as part of our IDS effort. The Oak Ridge National Laboratory DAAC will archive and disseminate this information as part of its 1997-8 activities through both a www-server and through CD-ROM. The National Snow and Ice Data Center DAAC is supporting the distribution of a pan-Arctic version of RivDIS v1.0 that will be supplemented with WMO-ACSYS, USGS, and Environment Canada data. We anticipate holdings for the Russian portion of the Arctic to increase to several hundred stations with an average spatial coverage of 104 - 105 km2.

"Pre-LBA" CD-ROM. We contributed a version of the STN-30 (v4.2) river networks for the Amazon/Tocantins River system to the ORNL-DAAC / IGBP-BAHC / INPA-CPTEC effort to develop a CD-ROM in anticipation of the upcoming Large-scale Biosphere-Atmosphere Experiment in Amazonia (LBA). The 30-minute spatial resolution networks link more than 2000 individual grid cells, as utilized in Vorosmarty et al. (1996b).

G. Participation in National and International Committees

Several activities involving both US and international organizations have been supported in part by our NASA EOS IDS grant.

Moore has Chaired the Global Analysis, Interpretation, and Modeling Task Force (GAIM) of the International Geosphere-Biosphere Programme (IGBP), and in January 1998, he will assume Chair of the Scientific Committee of the International Geosphere Biosphere Programme. Moore also chaired the Committee on Global Change Reserach of the National Academy of Sciences. W. R. Emanuel has served on the GAIM Task Force of the IGBP and on the National Technical Advisory Committee of the National Institute of Global Environmental Change.

Co-investigator Vorosmarty has served since 1993 as a member of the Scientific Steering Committee of the BAHC ("Biospheric Aspects of the Hydrological Cycle") Core Project, involved specifically with Project 8 Activities that examine the issues of anthropogenic change and drainage basin dynamics. He is a member in standing of the PAGES Fluvial Systems Working Group. He is also Secretary of the IAHS International Commission on Atmosphere-Soil-Vegetation Relations and began preparations for an IAHS Workshop on Global Hydrological Data Sets for the upcoming IAHS Birmingham Symposium in 1999.

Co-investigator Vorosmarty convened a workshop in December of 1994 on behalf of IGBP with the aim of developing a coherent strategy to address the issue of modeling river transports for both water and constituents from the continental land mass to the world's coastal oceans. Results of this workshop and recommendations for IGBP synthesis were published as an IGBP report (Vorosmarty et al. 1997d). Vorosmarty also participated as an invited keynote speaker on this subject in the Fourth IGBP-Scientific Advisory Committee meeting in Beijing (November 1995). The follow-up work performed in 1996 and 1997 is being published as a contribution to a Cambridge University Press book on Asian and Global Change (Vorosmarty et al. 1998b). A strategy paper written by IDS co-investigators Vorosmarty and Peterson (1998c) on fluvial transport is also being published as a contribution to the US-SCOPE Committee volume on estuarine synthesis.

W. R. Emanuel will participate in the GCTE-LUCC Open Science Conference on Global Change during March, 1998 in Barcelona, Spain. In collaboration with University of Virginia colleagues, he will present posters describing results from this project and related research.

New and Updated Publications and Reports

Ardo, J., Lambert, N., Henzlik, V., and Rock, B.N. 1997. Satellite-based estimations of coniferous forest cover changes in the Krusne Hory, Czech Republic, 1972-1989. Ambio, Vol. 26, no. 3, 158-166.

Asner, G.P., B.H. Braswell, D.S. Schimel, and C.A. Wessman. Ecological research needs from multi-angle remote sensing data. Remote Sensing of Environment, 1997.

Bondeau, A., J. Kaduk, D. Kicklighter, and the participants of the Potsdam NPP Model Intercomparison. Submitted. Comparing global models of terrestrial net primary productivity (NPP): Analysis of the seasonal behaviour of NPP, LAI, FPAR along climatic gradients across ecotones. Global Change Biology.

Braswell, B.H., D.S. Schimel, E. Linder, and B. Moore. The response of global terrestrial ecosystems to interannual temperature variability. Science, 1997.

Bubier, J.L., Rock, B.N. and Crill, P.M. 1997. Spectral reflectance measurements of boreal wetland and forest mosses. Journ. Geophys. Research, Vol. 102, No. D24, 29,483-29,494.

Cramer, W., D. W. Kicklighter, A. Bondeau, B. Moore III, G. Churkina, A. Ruimy, A. Schloss, and the participants of "POTSDAM '95". Submitted. Comparing global models of terrestrial net primary productivity (NPP): Overview and key results. Global Change Biology.

Dai, Z, C. Li, and R. Sass, 1997, (In preparation) Modeling methane emissions from rice paddies.

Field, C., D. Schimel, T. Ball, S. Cowling, R. Drapek, K. Hibbard, R. Kelly, D. Kicklighter, Y. Luo, G. Marion, D. McGuire, R. McKeown, R. McMurtrie, J. Melillo, R. Norby, W. Parton, A. Peterson, L. Pierce, L. Pitelka, A. Ruimy, S. Running, T. Smith, I. Woodward and D. Zak. In preparation. Reconciling modeling and experimental perspectives on ecosystem responses to elevated CO2: Role of N dynamics. BioScience.

Heimann, M., G. Esser, A. Haxeltine, J. Kaduk, D.W. Kicklighter, W. Knorr, G.H. Kohlmaier, A.D. McGuire, J. Melillo, B. Moore, R.D. Otto, I.C. Prentice, W. Sauf, A. Schloss, S. Sitch, U. Wittenberg and G. Wurth. In press. Evaluation of terrestrial carbon cycle models through simulations of the seasonal cycle of atmospheric CO2: First results of a model intercomparison study. Global Biogeochemical Cycles.

Holland, E.A., B.H. Braswell, A. Townsend, J.-F. Lamarque, J.-F. Muller, F. Dentener, G. Brasseur, H. Levy II, J.E. Penner, G. Roelofs, and J. Sulzman. The spatial distribution of nitrogen deposition and its impact on terrestrial ecosystems. Journal of Geophysical Research, 102(13):15849- 15966, 1997.

Jenkins, J.C., D.W. Kicklighter and J.D. Aber. Submitted. Predicting the regional impacts of increased CO2 and climatic change on forest productivity: A model comparison using PnET-II and TEM 4.0. IN: Responses of Northern U.S. Forests to Environmental Change, edited by R. Birdsey, J. Hom and R. Mickler.

Jenkins, J.C., D.W. Kicklighter, S.V. Ollinger, J.D. Aber and J.M. Melillo. In preparation. Sources of variability in NPP predictions at the regional scale: A comparison using PnET-II and TEM 4.0 in northeastern U.S. forests. Ecosystems.

Kicklighter, D.W., A. Bondeau, A.L. Schloss, J. Kaduk, A.D. McGuire and the other participants of "Potsdam '95". Submitted. Comparing global models of terrestrial net primary productivity (NPP): Global pattern and differentiation by major biomes. Global Change Biology.

Kicklighter, D.W., M. Bruno, S. Donges, G. Esser, M. Heimann, J. Helfrich, F. Ift, F. Joos, J. Kaduk, G.H. Kohlmaier, A.D. McGuire, J.M. Melillo, R. Meyer, B. Moore III, A. Nadler, I.C. Prentice, W. Sauf, A. Schloss, S. Sitch, U. Wittenberg and G. Wurth. Submitted. A first order analysis of the potential of CO2 fertilization to affect the global carbon budget: An intercomparison study of four terrestrial biosphere models. Tellus.

Lawless, J.G. and Rock, B.N. 1998 (in press). Student Scientist Partnerships and data quality. Journ. Science Educ. and Technology.

Li, C., J. Aber, F. Stange, K. Butterbach-Bahl, and H. Papen. In preparation. A model of nitrous oxide and nitric oxide emissions from forest ecosystems: 1. Model development.

Li, C., M. Keller, P. Crill, E. Veldcamp, A. Weltz, In preparation. Nitrous Oxide Emissions from Agricultural Soils in Costa Rica: 2. Model Simulation.

McGuire, A.D., and J.E. Hobbie. 1997. Global climate change and the equilibrium responses of carbon storage in arctic and subarctic regions. pp. 47-48. IN: Arctic System Science Modeling Workshop Report. The Arctic Research Consortium of the United States. Fairbanks, Alaska.

McGuire, A.D., J.E. Hobbie, D.W. Kicklighter, B.L. Kwaitkowski, J. Helfrich and E.B. Rastetter. In preparation. Carbon storage responses of tundra ecosystems to climatic variation: Retrospective and future assessments of fine- and coarse-scale models at different spatial scales. Global Change Biology.

McGuire, A.D., D.W. Kicklighter, and J.M. Melillo. 1996. Global climate change and carbon cycling in grasslands and conifer forests. pp. 389-411. IN: Global Change: Effects on Coniferous Forests and Grasslands, SCOPE Volume 56. Edited by A.I. Breymeyer, D.O. Hall, J.M. Melillo and G.I. Agren. John Wiley and Sons, Chichester, United Kingdom.

McGuire, A.D., J.M. Melillo, D.W. Kicklighter, Y. Pan, X. Xiao, J. Helfrich, B. Moore III, C.J. Vorosmarty, and A.L. Schloss. 1997. Equilibrium responses of global net primary production and carbon storage to doubled atmospheric carbon dioxide: Sensitivity to changes in vegetation nitrogen concentration. Global Biogeochemical Cycles 11: 173-189.

Melillo, J.M., R.A. Houghton, D.W. Kicklighter, and A.D. McGuire. 1996. Tropical deforestation and the global carbon budget. Annual Review of Energy and the Environment 21: 293-310.

Melillo, J.M., D.W. Kicklighter, J. Helfrich, A.D. McGuire, B. Moore III, C.J. Vorosmarty, and.A.L. Schloss In preparation. The effect of CO2 fertilization on the storage of carbon in terrestrial ecosystems: A global modeling study.

Moss, D.M., Rock, B.N., Bogle, A.L., and Bilkova, J. 1997 (in press). Anatomical evidence of the development of damage symptoms across a growing season in needles of red spruce from central New Hampshire. Environ. Experimental Bot.

Pan, Y., A.D. McGuire, D.W. Kicklighter, and J.M. Melillo. 1996. The importance of climate and soils on estimates of net primary production: A sensitivity analysis with the Terrestrial Ecosystem Model. Global Change Biology 2: 5-23.

Pan, Y., J.M. Melillo, D.W. Kicklighter, X. Xiao, and A.D. McGuire. In preparation. Potential response of net primary productivity in terrestrial ecosystems of China to climate changes: A simulation study by Terrestrial Ecosystem Model coupled with vegetation redistribution.

Pan, Y., J.M. Melillo, A.D. McGuire, D.W. Kicklighter, L.F. Pitelka, K. Hibbard, L.L. Pierce, S.W. Running, D.S. Ojima, W.J. Parton, D.S. Schimel and other VEMAP Members. In press. Response of terrestrial ecosystems to elevated atmospheric CO2: A comparison of simulation studies among biogeochemistry models. Oecologia.

Perez-Garcia, J., L.A. Joyce, C.S. Binkley, and A.D. McGuire. In press. Economic impacts of climatic change on the global forest sector: an integrated ecological/economic assessment. Critical Reviews in Science and Technology.

Prinn, R., H. Jacoby, A. Sokolov, C. Wang, X. Xiao, Z. Yang, R. Eckaus, P. Stone, D. Ellerman, J. Melillo, J. Fitzmaurice, D. Kicklighter, G. Holian and Y. Liu. Submitted. Integrated global system model for climate policy assessment: Feedbacks and sensitivity studies. Climatic Change.

Rock, B.N., Blackwell, T.R., Miller, D., and Hardison, A. 1997. The GLOBE Program: A Model for International Education. In: Internet Links for Science Education: Student-Scientist Partnerships, Cohen, K., Editor, Plenum Publ. Corp. NY, p. 17-30.

Rock, B.N. and Lawless, J.G. 1997. The GLOBE Program: A source of datasets for use in global change studies. IGBP Newsletter 29, 15-17.

Rock, B.N. and Lauten, G.N. 1996. K-12th grade students as active contributors to research investigations. Journ. Science Educ. and Technology, 5: 255-266.

Schimel, D.S., B.H. Braswell, and W.J. Parton. Equilibration dynamics of the terrestrial water, nitrogen, and carbon cycles: conclusions from a preindustrial simulation. Proceedings of the National Academy of Sciences, 94:8280-8283, 1997.

Schimel, D.S., VEMAP Participants, and B.H. Brasswell. 1997. Continental scale variability in ecosystem processes: Models, data and the role of disturbance. Ecological Monographs 67: 251-271.

Schloss, A.L., U. Wittenberg, D.W. Kicklighter, J. Kaduk and the other participants of "Potsdam '95". Submitted. Comparing global models of terrestrial net primary productivity (NPP): Relationships of annual NPP to spatial climatic drivers and the normalized difference vegetation index. Global Change Biology.

Sharma, K. 1997. Impact of Land-Use and Climatic Changeson Hydrology of the Himalayan Basin: A Case Study in the KosiBasin. PhD Dissertation. Earth Science Dept., Universityof New Hampshire, Durham.

Stange, F., K. Butterbach-Bahl, H. Papen, Li, C., and J. Aber, In preparation. A model of nitrous oxide and nitric oxide emissions from forest ecosystems: 2. Model applications

Steudler, P.A., J.M. Melillo, B.J. Feigl, C. Neill, M.C. Piccolo, and C. C. Cerri. 1996. Consequence of forest-to-pasture conversion on CH4 fluxes in the Brazilian Amazon Basin. Journal of Geophysical Research 101: 18,547-18,554.

Tian, H., J.M. Melillo, D.W. Kicklighter and A.D. McGuire. Submitted. The sensitivity of terrestrial carbon storage to historical atmospheric CO2 and climate variability in the United States. Ecological Applications.

Tian, H., J.M. Melillo, D.W. Kicklighter, A.D. McGuire, B. Moore III and C.J. Vorosmarty. In preparation. Influence of climate variability and CO2 on carbon storage in undisturbed ecosystems in the Amazon Basin. Nature.

Vorosmarty, C.J., C.A. Federer and A. Schloss. 1998a. Potential evaporation functions compared on U.S. watersheds:Implications for global-scale water balance and terrestrial ecosystemmodeling. In press: J. of Hydrology.

Vorosmarty, C.J., C. Li, J. Sun, and Z. Dai. 1998b. Emergingimpacts of anthropogenic change on global river systems: TheChinese example. In: J. Galloway and J. Melillo (eds.), Asian Change in the Context of Global Change: Impacts of Naturaland Anthropogenic Changes in Asia on Global Biogeochemical Cycles. Cambridge: Cambridge University Press. In press.

Vorosmarty, C.J. and B.J. Peterson. 1998c. Macro-scalemodels of water and nutrient flux to the coastal zone. In: J.Hobbie (ed)., SCOPE Estuarine Synthesis. In review.

Vorosmarty, C.J. Sharma, K., Fekete, B., Copeland, A.H.,Holden, J., Marble, J. and J.A. Lough. 1997a. The storage andaging of continental runoff in large reservoir systems of theworld. In press: Ambio.

Vorosmarty, C.J., M. Meybeck, B. Fekete, and K. Sharma.1997c. The potential impact of neo-Castorization on sedimenttransport by the global network of rivers. In: D. Walling (ed.),Human Impact on Erosion and Sedimentation. IAHS Press,Wallingford UK.

Vorosmarty, C.J., R. Wasson, and J.E. Richey (eds.). 1997d. Modeling the Transport and Transformation of TerrestrialMaterials to Freshwater and Coastal Ecosystems. WorkshopReport and Recommendations for IGBP Inter-Core Project Collaboration. IGBP Secretariat, Stockholm. (In press).

Xiao, X., D.W. Kicklighter, J.M. Melillo, A.D. McGuire, P.H. Stone, and A.P. Sokolov. 1997. Linking a global terrestrial biogeochemical model and a 2 dimensional climate model: implications for the carbon budget. Tellus 49B: 18-37.

Xiao, X., J.M. Melillo, D.W. Kicklighter, A.D. McGuire, P.H. Stone, and A.P. Sokolov. 1996. Relative roles of changes in CO2 and climate to equilibrium responses of net primary production and carbon storage of the terrestrial biosphere. MIT Joint Program on Science and Policy of Global Change Report No. 8. Massachusetts Institute of Technology, Cambridge, Massachusetts. 34 p. ( see also http://web.mit.edu/globalchange/www/rpt8a.html)

Xiao, X., J.M. Melillo, D.W. Kicklighter, A.D. McGuire, R.G. Prinn, C. Wang, P.H. Stone and A. Sokolov. Submitted. Transient climate change and net ecosystem production of the terrestrial biosphere. Global Biogeochemical Cycles.

Xiao, X., J.M. Melillo, D.W. Kicklighter, A.D. McGuire, H. Tian, Y. Pan, C.J. Vorosmarty and Z. Yang. 1997. Submitted. Transient climate change and potential croplands of the world in the 21st century. Ambio.

Xiao, X., J.M. Melillo, D.W. Kicklighter, Y. Pan, A.D. McGuire, and J. Helfrich. In press. Net primary production of terrestrial ecosystems in China and its equilibrium responses to changes in climate and atmospheric CO2 concentration. Acta Phytoecologia Sinica.

Additional References

Ball, J. T., I. E. Woodrow, and J. A. Berry. 1987. A model predicting stomatal conductance and its contribution to the control of photosynthesis under different environmental conditions. pp. 221-224. In I. Biggins (ed.) Progress in Photosynthesis Research Vol. IV Proceedings of the VIIth International Congress on Photosynthesis. Nijhoff, Dordrecht, Netherlands.

Emanuel, W. R., A. W. King, and W. M. Post. 1994. A dynamic model of terrestrial carbon cycling. pp. 239--260. In M. Heimann (ed.) The Global Carbon Cycle. Springer-Verlag, Berlin.

Emanuel, W. R. 1996. Modeling carbon cycling on disturbed landscapes. Ecological Modelling 89:1--12.

Friend, A. D. 1991. Use of a model of photosynthesis and leaf microenvironment to predict optimal stomatal conductance and leaf nitrogen partitioning. Plant, Cell and Environment 14:895--905.

Monteith, J. L. 1995. A reinterpretation of stomatal responses to humidity. Plant, Cell and Environment 18:357--364.

Parton, W. J., J. M. O. Scurlock, D. S. Ojima, T. G. Gilmanov, R. J. Scholes, D. S. Schimel, T. Kirchner, J-C. Menaut, T. Seastedt, E. Garcia Moya, A. Kamnalrut, and J. I. Kinyamario. 1993. Observations and modeling of biomass and soil organic matter dynamics for the grassland biome worldwide. Global Biogeochemical Cycles 7:785--809.

Raich, J. W., E. B. Rastetter, J. M. Melillo, D. W. Kicklighter, P. A. Steudler, B. J. Peterson, A. L. Grace, B. Moore III, and C. J. Vorosmarty. 1991. Potential net primary productivity in South America: Application of a global model. Ecological Applications 1:359--429.

Woodward, F. I., T. M. Smith, and W. R. Emanuel. 1995. A global primary productivity and phytogeography model. Global Biogeochemical Cycles 9:417--490.