Numerical models based on physics represent the state of the art in Earth system modeling and comprise our best tools for generating insights and predictions. Despite rapid growth in computational power, the perceived need for higher model resolutions overwhelms the latest generation computers, reducing the ability of modelers to generate simulations for understanding parameter sensitivities and characterizing variability and uncertainty. Thus, surrogate models are often developed to capture the essential attributes of the full-blown numerical models. Recent successes of machine learning methods, especially deep learning (DL), across many disciplines offer the possibility that complex nonlinear connectionist representations may be able to capture the underlying complex structures and nonlinear processes in Earth systems. A difficult test for DL-based emulation, which refers to function approximation of numerical models, is to understand whether they can be comparable to traditional forms of surrogate models in terms of computational efficiency while simultaneously reproducing model results in a credible manner. A DL emulation that passes this test may be expected to perform even better than simple models with respect to capturing complex processes and spatiotemporal dependencies. Here, we examine, with a case study in satellite-based remote sensing, the hypothesis that DL approaches can credibly represent the simulations from a surrogate model with comparable computational efficiency. Our results are encouraging in that the DL emulation reproduces the results with acceptable accuracy and often even faster performance. We discuss the broader implications of our results in light of the pace of improvements in high-performance implementations of DL and the growing desire for higher resolution simulations in the Earth sciences.

Applications of satellite data in areas such as weather tracking and modeling, ecosystem monitoring, wildfire detection, and land-cover change are heavily dependent on the tradeoffs to spatial, spectral, and temporal resolutions of observations. In weather tracking, high-frequency temporal observations are critical and used to improve forecasts, study severe events, and extract atmospheric motion, among others. However, while the current generation of geostationary (GEO) satellites has hemispheric coverage at 10-15-min intervals, higher temporal frequency observations are ideal for studying mesoscale severe weather events. In this work, we present a novel application of deep learning-based optical flow to temporal upsampling of GEO satellite imagery. We apply this technique to 16 bands of the GOES-R/Advanced Baseline Imager mesoscale dataset to temporally enhance full-disk hemispheric snapshots of different spatial resolutions from 10 to 1 min. Experiments show the effectiveness of task-specific optical flow and multiscale blocks for interpolating high-frequency severe weather events relative to bilinear and global optical flow baselines. Finally, we demonstrate strong performance in capturing variability during convective precipitation events.

Atmospheric winds are a key physical phenomenon impacting natural hazards, energy transport, ocean currents, large-scale circulation, and ecosystem fluxes. Observing winds is a complex process and presents a large gap in NASA's Earth Observation System. Atmospheric motion vectors (AMVs) aim to fill this gap by making numerical estimates of cloud movement between sequences of multi-spectral satellite images, tracking clouds and water vapor. Recent imaging hardware and software advancements have enabled the use of numerical optical flow techniques to produce accurate and dense vector fields outperforming traditional methods. This work presents WindFlow as the first machine learning based system for feature tracking atmospheric motion using optical flow. Due to the lack of large-scale satellite-based observations, we leverage high-resolution numerical simulations from NASA's GEOS-5 Nature Run to perform supervised learning and transfer to satellite images. We demonstrate that our approach using deep learning based optical flow scales to ultra-high-resolution images of size 2881x5760 with less than 1 m/s bias and 2.5 m/s average error. Four network and learning architectures are compared and it is found that recurrent all-pairs field transforms (RAFT) produces the lowest errors on all metrics for wind speed and direction. Results on held out numerical outputs show RAFT's good performance in each of the spatial, temporal, and physical dimensions. A comparison between WindFlow and an operational AMV product against rawinsonde observations shows that RAFT transfers across simulations and thermal infrared satellite observations. This work shows that machine learning based optical flow is an efficient approach to generating robust feature tracking for AMVs consistently over large regions.

Earth-observing satellites carrying multispectral sensors are widely used to monitor the physical and biological states of the atmosphere, land, and oceans. These satellites have different vantage points above the Earth and different spectral imaging bands resulting in inconsistent imagery from one to another. This presents challenges in building downstream applications. What if we could generate synthetic bands for existing satellites from the union of all domains? We tackle the problem of generating synthetic spectral imagery for multispectral sensors as an unsupervised image-to-image translation problem modeled with a variational autoencoder (VAE) and generative adversarial network (GAN) architecture. Our approach introduces a novel shared spectral reconstruction loss to constrain the high-dimensional feature space of multispectral images. Simulated experiments performed by dropping one or more spectral bands show that cross-domain reconstruction outperforms measurements obtained from a second vantage point. Our proposed approach enables the synchronization of multispectral data and provides a basis for more homogeneous remote sensing datasets.

We extended the MODIS 17 algorithm to hourly Gross Primary Product (GPP) estimation for Australia in 2018. AHI 1-km data are available at NASA Earth Exchange (NEX) facility. The high-frequent observation of Advanced Himawari Imager (AHI) enabled us to calculate hourly GPP, which is corresponding to ground observation of carbon flux measured by TERN OzFlux. We interpolated the weather observation data to 1-km grid over Australia ingesting AHI reflectance and radiometric temperature data (NEX-Gridded Hourly Meteorology (NEX-GHM)). Then, we applied MODIS 17 algorithm to each grid with optimized climate regulation functions. The estimation of climate and GPP were well correlated with OzFlux ground observation data.

The latest generation of geostationary satellites (Himawari 8/9, GOES-16/17, FY-4, GK-2A) carries sensors that closely mimic the spatial and spectral characteristics of widely used polar-orbiting, global monitoring sensors such as MODIS and VIIRS. When combined, data from various currently operating/planned geostationary platforms provide a geo-ring of hyper-temporal (5-10 minutes), multispectral observations at spatial resolutions as high as 500 m. These high frequency observations offer exciting new possibilities for monitoring our planet, including better retrievals of geophysical variables by overcoming cloud cover, enabling studies of diurnally varying phenomena in the atmosphere, land, and the oceans, and support operational decision-making in agriculture, hydrology and disaster management. The NASA Earth Exchange (NEX) team, in collaboration with scientists from JAXA, KARI, NOAA and other international institutions, created the GeoNEX (www.nasa.gov/geonex) pipeline to integrate data from all available geostationary platforms and produce and distribute spatially, temporally, and radiometrically consistent data for the earth science community. We envision various institutions adapting the Geo component (e.g., GeoNOAA, GeoKARI, GeoChiba, GeoJAXA, GeoCMA) and customizing the pipeline and downstream products to serve the local/regional research and applied science communities.

Recent progress in quantum algorithms and hardware indicates the potential importance of quantum computing in the near future. However, finding suitable application areas remains an active area of research. Quantum machine learning is touted as a potential approach to demonstrate quantum advantage within both the gate-model and the adiabatic schemes. For instance, the Quantum-assisted Variational Autoencoder (QVAE) has been proposed as a quantum enhancement to the discrete VAE. We extend on previous work and study the real-world applicability of a QVAE by presenting a proof-of-concept for similarity search in large-scale high-dimensional datasets. While exact and fast similarity search algorithms are available for low dimensional datasets, scaling to high-dimensional data is non-trivial. We show how to construct a space-efficient search index based on the latent space representation of a QVAE. Our experiments show a correlation between the Hamming distance in the embedded space and the Euclidean distance in the original space on the Moderate Resolution Imaging Spectroradiometer (MODIS) dataset.Further, we find real-world speedups compared to linear search and demonstrate memory-efficient scaling to half a billion data points.

The latest generation of geostationary satellites carry sensors such as the Advanced Baseline Imager (GOES-16/17) and the Advanced Himawari Imager (Himawari-8/9) that closely mimic the spatial and spectral characteristics of widely used polar orbiting sensors such as EOS/MODIS. More importantly, they provide observations at 1-5-15 minute intervals, instead of twice a day from MODIS, offering unprecedented opportunities for monitoring large parts of the Earth. In addition to serving the needs of weather forecasting, these observations offer new and exciting opportunities in managing solar power, fighting wildfires, and tracking air pollution. Creation of actionable information in near realtime from these data streams is a challenge that is best addressed through collaborative efforts among the industry, academia and government agencies.

The impacts of climate change are felt by most critical systems, such as infrastructure, ecological systems, and power-plants. However, contemporary Earth System Models (ESM) are run at spatial resolutions too coarse for assessing effects this localized. Local scale projections can be obtained using statistical downscaling, a technique which uses historical climate observations to learn a low-resolution to high-resolution mapping. The spatio-temporal nature of the climate system motivates the adaptation of super-resolution image processing techniques to statistical downscaling. In our work, we present DeepSD, a generalized stacked super resolution convolutional neural network (SRCNN) framework with multi-scale input channels for statistical downscaling of climate variables. A comparison of DeepSD to four state-of-the-art methods downscaling daily precipitation from 1 degree (~100km) to 1/8 degrees (~12.5km) over the Continental United States. Furthermore, a framework using the NASA Earth Exchange (NEX) platform is discussed for downscaling more than 20 ESM models with multiple emission scenarios.

Deep Learning (DL) methods have been transforming computer vision with innovative adaptations to other domains including climate change. For DL to pervade Science and Engineering (S&EE) applications where risk management is a core component, well-characterized uncertainty estimates must accompany predictions. However, S&E observations and model-simulations often follow heavily skewed distributions and are not well modeled with DL approaches, since they usually optimize a Gaussian, or Euclidean, likelihood loss. Recent developments in Bayesian Deep Learning (BDL), which attempts to capture uncertainties from noisy observations, aleatoric, and from unknown model parameters, epistemic, provide us a foundation. Here we present a discrete-continuous BDL model with Gaussian and lognormal likelihoods for uncertainty quantification (UQ). We demonstrate the approach by developing UQ estimates on "DeepSD'', a super-resolution based DL model for Statistical Downscaling (SD) in climate applied to precipitation, which follows an extremely skewed distribution. We find that the discrete-continuous models outperform a basic Gaussian distribution in terms of predictive accuracy and uncertainty calibration. Furthermore, we find that the lognormal distribution, which can handle skewed distributions, produces quality uncertainty estimates at the extremes. Such results may be important across S&E, as well as other domains such as finance and economics, where extremes are often of significant interest. Furthermore, to our knowledge, this is the first UQ model in SD where both aleatoric and epistemic uncertainties are characterized.

The impacts of climate change are felt by most critical systems, such as infrastructure, ecological systems, and power-plants. However, contemporary Earth System Models (ESM) are run at spatial resolutions too coarse for assessing effects this localized. Local scale projections can be obtained using statistical downscaling, a technique which uses historical climate observations to learn a low-resolution to high-resolution mapping. Depending on statistical modeling choices, downscaled projections have been shown to vary significantly terms of accuracy and reliability. The spatio-temporal nature of the climate system motivates the adaptation of super-resolution image processing techniques to statistical downscaling. In our work, we present DeepSD, a generalized stacked super resolution convolutional neural network (SRCNN) framework for statistical downscaling of climate variables. DeepSD augments SRCNN with multi-scale input channels to maximize predictability in statistical downscaling. We provide a comparison with Bias Correction Spatial Disaggregation as well as three Automated-Statistical Downscaling approaches in downscaling daily precipitation from 1 degree (~100km) to 1/8 degrees (~12.5km) over the Continental United States. Furthermore, a framework using the NASA Earth Exchange (NEX) platform is discussed for downscaling more than 20 ESM models with multiple emission scenarios.

Many real-world problems involve learning models for rare classes in situations where there are no gold standard labels for training samples but imperfect labels are available for all instances. In this paper, we present RAPT, a three step predictive modeling framework for classifying rare class in such problem settings. The first step of the proposed framework learns a classifier that jointly optimizes precision and recall by only using imperfectly labeled training samples. We also show that, under certain assumptions on the imperfect labels, the quality of this classifier is almost as good as the one constructed using perfect labels. The second and third steps of the framework make use of the fact that imperfect labels are available for all instances to further improve the precision and recall of the rare class. We evaluate the RAPT frameworkon two real-world applications of mapping forest fires and urban extent from earth observing satellite data. The experimental results indicate that RAPTcan be used to identifyforest fires and urban areas with high precision and recall by using imperfect labels, even though obtaining expert annotated samples on a global scale is infeasible in these applications.

We investigate the use of Deep Neural Networks for the classification of image datasets where texture features are important for generating class-conditional discriminative representations. To this end, we first derive the size of the feature space for some standard textural features extracted from the input dataset and then use the theory of Vapnik-Chervonenkis dimension to show that hand-crafted feature extraction creates low-dimensional representations which help in reducing the overall excess error rate. As a corollary to this analysis, we derive for the first time upper bounds on the VC dimension of Convolutional Neural Network as well as Dropout and Dropconnect networks and the relation between excess error rate of Dropout and Dropconnect networks. The concept of intrinsic dimension is used to validate the intuition that texture-based datasets are inherently higher dimensional as compared to handwritten digits or other object recognition datasets and hence more difficult to be shattered by neural networks. We then derive the mean distance from the centroid to the nearest and farthest sampling points in an n-dimensional manifold and show that the Relative Contrast of the sample data vanishes as dimensionality of the underlying vector space tends to infinity.

Satellite image classification is a challenging problem that lies at the crossroads of remote sensing, computer vision, and machine learning. Due to the high variability inherent in satellite data, most of the current object classification approaches are not suitable for handling satellite datasets. The progress of satellite image analytics has also been inhibited by the lack of a single labeled high-resolution dataset with multiple class labels. The contributions of this paper are twofold -- (1) first, we present two new satellite datasets called SAT-4 and SAT-6, and (2) then, we propose a classification framework that extracts features from an input image, normalizes them and feeds the normalized feature vectors to a Deep Belief Network for classification. On the SAT-4 dataset, our best network produces a classification accuracy of 97.95% and outperforms three state-of-the-art object recognition algorithms, namely - Deep Belief Networks, Convolutional Neural Networks and Stacked Denoising Autoencoders by ~11%. On SAT-6, it produces a classification accuracy of 93.9% and outperforms the other algorithms by ~15%. Comparative studies with a Random Forest classifier show the advantage of an unsupervised learning approach over traditional supervised learning techniques. A statistical analysis based on Distribution Separability Criterion and Intrinsic Dimensionality Estimation substantiates the effectiveness of our approach in learning better representations for satellite imagery.

Accurate tree-cover estimates are useful in deriving above-ground biomass density estimates from very high resolution (VHR) satellite imagery data. Numerous algorithms have been designed to perform tree-cover delineation in high-to-coarse-resolution satellite imagery, but most of them do not scale to terabytes of data, typical in these VHR data sets. In this paper, we present an automated probabilistic framework for the segmentation and classification of 1-m VHR data as obtained from the National Agriculture Imagery Program (NAIP) for deriving tree-cover estimates for the whole of Continental United States, using a high-performance computing architecture. The results from the classification and segmentation algorithms are then consolidated into a structured prediction framework using a discriminative undirected probabilistic graphical model based on conditional random field, which helps in capturing the higher order contextual dependence relations between neighboring pixels. Once the final probability maps are generated, the framework is updated and retrained by incorporating expert knowledge through the relabeling of misclassified image patches. This leads to a significant improvement in the true positive rates and reduction in false positive rates (FPRs). The tree-cover maps were generated for the state of California, which covers a total of 11 095 NAIP tiles and spans a total geographical area of 163 696 sq. miles. Our framework produced correct detection rates of around 88% for fragmented forests and 74% for urban tree-cover areas, with FPRs lower than 2% for both regions. Comparative studies with the National Land-Cover Data algorithm and the LiDAR high-resolution canopy height model showed the effectiveness of our algorithm for generating accurate high-resolution tree-cover maps.

NASA Earth Exchange (NEX) is a collaboration platform whose goal is to accelerate Earth science research, by leveraging NASA's vast collections of global satellite data together with access to NASA's High-End Computing (HEC) facilities. NEX also aims to facilitate the sharing of experimental results as well as scientific processes (workflows) with the Earth science community through integration with VisTrails workflow management system. While VisTrails is used internally, it is not easily accessible from remote computers without directly logging into the NASA HEC systems through twofactor authentication and a bastion host. This paper describes the initial design of an extensible architecture that facilitates easier workflow interaction on NEX, by enabling users to develop and execute workflows in a supercomputing environment directly from their local VisTrails installation. This architecture helps domain scientists seamlessly leverage distributed computing and storage resources and it is potentially applicable to other scientific workflow management software. We further describe the architecture of the VisTrails-HEC plugin (as well as the VisTrails-Amazon plugin) and the implementation of a working prototype to demonstrate the feasibility of our solution.

The gradients of the Appalachian Trail (A.T.) in elevations and latitudes provide a megatransect to study environmental variations in the eastern United States. This paper reveals patterns and trends of land surface phenology (LSP) in association with climatic variables within a corridor area along the A.T. We employed time-series data from Global Inventory Modeling and Mapping Studies and the Surface Observation and Gridding System between 1982 and 2006 to extract spatial and temporal variation patterns of LSP metrics and the correlations with meteorological parameters. The derived trends in LSP metrics indicate that the extended length of season mainly resulted from delayed end of season (EOS) across the study area. More significant change occurred in the northern segment than in the southern segment, which reflects latitudinal effects. We analyzed the relationship between LSP and longitude, latitude, elevation, local climatic variables, and large-scale climate oscillations. Delayed start of season in 1989 and advanced EOS in 1988 were observed responding to the La Niña episode during 1988–1989. This paper provides information about the effects of climate and topography on LSP along the Appalachian Mountain ridges.

This paper introduces a collaborative multi-agency effort to develop an Appalachian Trail (A.T.) MEGA-Transect Decision Support System (DSS) for monitoring, reporting and forecasting ecological conditions of the A.T. and the surrounding lands. The project is to improve decisionmaking on management of the A.T. by providing a coherent framework for data integration, status reporting and trend analysis. The A.T. MEGA-Transect DSS is to integrate NASA multi-platform sensor data and modeling through the Terrestrial Observation and Prediction System (TOPS) and in situ measurements from A.T. MEGA-Transect partners to address identified natural resource priorities and improve resource management decisions.

The Moderate Resolution Spectroradiometer (MODIS) sensor has provided near real-time estimates of gross primary production (GPP) since March 2000. We compare four years (2000 to 2003) of satellite-based calculations of GPP with tower eddy CO2flux-based estimates across diverse land cover types and climate regimes. We examine the potential error contributions from meteorology, leaf area index (LAI)/fPAR, and land cover. The error between annual GPP computed from NASA's Data Assimilation Office's (DAO) and tower-based meteorology is 28%, indicating that NASA's DAO global meteorology plays an important role in the accuracy of the GPP algorithm. Approximately 62% of MOD15-based estimates of LAI were within the estimates based on field optical measurements, although remaining values overestimated site values. Land cover presented the fewest errors, with most errors within the forest classes, reducing potential error. Tower-based and MODIS estimates of annual GPP compare favorably for most biomes, although MODIS GPP overestimates tower-based calculations by 20%-30%. Seasonally, summer estimates of MODIS GPP are closest to tower data, and spring estimates are the worst, most likely the result of the relatively rapid onset of leaf-out. The results of this study indicate, however, that the current MODIS GPP algorithm shows reasonable spatial patterns and temporal variability across a diverse range of biomes and climate regimes. So, while continued efforts are needed to isolate particular problems in specific biomes, we are optimistic about the general quality of these data, and continuation of the MOD17 GPP product will likely provide a key component of global terrestrial ecosystem analysis, providing continuous weekly measurements of global vegetation production

The analysis of two years of Collection 3 and five years of Collection 4 Terra Moderate Resolution Imaging Spectroradiometer (MODIS) Leaf Area Index (LAI) and Fraction of Photosynthetically Active Radiation (FPAR) data sets is presented in this article with the goal of understanding product quality with respect to version (Collection 3 versus 4), algorithm (main versus backup), snow (snow-free versus snow on the ground), and cloud (cloud-free versus cloudy) conditions. Retrievals from the main radiative transfer algorithm increased from 55% in Collection 3 to 67% in Collection 4 due to algorithm refinements and improved inputs. Anomalously high LAI/FPAR values observed in Collection 3 product in some vegetation types were corrected in Collection 4. The problem of reflectance saturation and too few main algorithm retrievals in broadleaf forests persisted in Collection 4. The spurious seasonality in needleleaf LAI/FPAR fields was traced to fewer reliable input data and retrievals during the boreal winter period. About 97% of the snow covered pixels were processed by the backup Normalized Difference Vegetation Index-based algorithm. Similarly, a majority of retrievals under cloudy conditions were obtained from the backup algorithm. For these reasons, the users are advised to consult the quality flags accompanying the LAI and FPAR product.

Global products of vegetation green Leaf Area Index (LAI) and Fraction of Photosynthetically Active Radiation absorbed by vegetation (FPAR) are being operationally produced from Terra and Aqua Moderate Resolution Imaging Spectroradiometers (MODIS) at 1-km resolution and eight-day frequency. This paper summarizes the experience of several collaborating investigators on validation of MODIS LAI products and demonstrates the close connection between product validation and algorithm refinement activities. The validation of moderate resolution LAI products includes three steps: 1) field sampling representative of LAI spatial distribution and dynamic range within each major land cover type at the validation site; 2) development of a transfer function between field LAI measurements and high resolution satellite data to generate a reference LAI map over an extended area; and 3) comparison of MODIS LAI with aggregated reference LAI map at patch (multipixel) scale in view of geo-location and pixel shift uncertainties. The MODIS LAI validation experiences, summarized here, suggest three key factors that influence the accuracy of LAI retrievals: 1) uncertainties in input land cover data, 2) uncertainties in input surface reflectances, and 3) uncertainties from the model used to build the look-up tables accompanying the algorithm. This strategy of validation efforts guiding algorithm refinements has led to progressively more accurate LAI products from the MODIS sensors aboard NASA's Terra and Aqua platforms

Application of remote sensing data to extrapolate evapotranspiration (ET) measured at eddy covariance flux towers is a potentially powerful method to estimate continental-scale ET. In support of this concept, we used meteorological and flux data from the AmeriFlux network and an inductive machine learning technique called support vector machine (SVM) to develop a predictive ET model. The model was then applied to the conterminous U.S. In this process, we first trained the SVM to predict 2000–2002 ET measurements from 25 AmeriFlux sites using three remotely sensed variables [land surface temperature, enhanced vegetation index (EVI), and land cover] and one ground-measured variable (surface shortwave radiation). Second, we evaluated the model performance by predicting ET for 19 flux sites in 2003. In this independent evaluation, the SVM predicted ET with a root-mean-square error (rmse) of 0.62 mm/day (approximately 23% of the mean observed values) and an$R^2$of 0.75. The rmse from SVM was significantly smaller than that from neural network and multiple-regression approaches in a cross-validation experiment. Among the explanatory variables, EVI was the most important factor. Indeed, removing this variable induced an rmse increase from 0.54 to 0.77 mm/day. Third, with forcings from remote sensing data alone, we used the SVM model to predict the spatial and temporal distributions of ET for the conterminous U.S. for 2004. The SVM model captured the spatial and temporal variations of ET at a continental scale.

Automated cloud detection and tracking is an important step in assessing changes in radiation budgets associated with global climate change via remote sensing. Data products based on satellite imagery are available to the scientific community for studying trends in the Earth's atmosphere. The data products include pixel-based cloud masks that assign cloud-cover classifications to pixels. Many cloud-mask algorithms have the form of decision trees. The decision trees employ sequential tests that scientists designed based on empirical astrophysics studies and simulations. Limitations of existing cloud masks restrict our ability to accurately track changes in cloud patterns over time. In a previous study we compared automatically learned decision trees to cloud masks included in Advanced Very High Resolution Radiometer (AVHRR) data products from the year 2000. In this paper we report the replication of the study for five-year data, and for a gold standard based on surface observations performed by scientists at weather stations in the British Islands. For our sample data, the accuracy of automatically learned decision trees was greater than the accuracy of the cloud masks p < 0.001.

Terrestrial evapotranspiration is a critical process for the energy and hydrologic partitioning at the land surface. We have developed an estimation method of the evaporation fraction, the energy budget equivalent of the ratio of actual to potential evapotranspiration, for use with the Aqua/MODIS sensor. The method normalizes the radiometric surface temperature with normalized difference vegetation index to infer changes in the land surface sensible/latent energy partitioning, and is computed every eight days globally. We tested the algorithm using Terra/MODIS datasets from the year 2000. Potential applications for the evaporation fraction include water resources management, mapping vegetation stress and wildland fire risk.

The Southeastern United States has undergone one of the highest rates of landscape changes in the country, as a result of changing demographics and land use practices over the last few decades. Here, we propose a remote sensing based methodology to estimate regional impacts of land development on ecosystem resources. As an indicator of ecosystem functioning we chose net primary productivity (NPP), which is routinely estimated from the MODerate resolution Imaging Spectroradiometer (MODIS) data. We combined a potential natural vegetation map, a current LANDSAT-based land cover map, nighttime data derived from the Defense Meteorological Satellite Program's Operational Linescan System (DMSP/OLS) combined in a geographic information system. We compared the effects of historical and recent land cover changes on NPP. The analysis indicates that 80% of the original land cover of the region has undergone changes that resulted in an annual reduction in NPP of 9%. In 1992, urban areas amounted to only 4% of the total surface in the region, but they appear to be responsible for most of the landscape fragmentation. Estimates based on nightlights data between 1992-93 and 2000 showed developed area increased by another 0.5%, contributing to a further reduction in annual NPP. The combination of MODIS products like NPP with nighttime data could provide rapid assessment of land cover changes and their impacts on regional ecosystem resources.

Satellite data are widely used in land surface models to compute carbon and water exchange processes. However, much of this work is retrospective in nature. To better represent current land surface conditions in weather/climate models or to provide timely information on ecosystem conditions for natural resource management, one must move from retrospective to real-time analysis. A number of advances allow us to develop a system that would allow such real-time assimilation. These include consistent and timely availability of land surface products from EOS/MODIS, and on-line availability of weather data from a number or surface weather stations. We have developed a data assimilation system, terrestrial observation and prediction system, that integrates satellite data, surface weather observations and weather/climate forecasts with a terrestrial ecosystem model. TOPS produces daily 1 km estimates of carbon and water fluxes using MODIS derived LAI, land cover and gridded meteorological data created using more than 2000 surface weather stations over the conterminous U.S. Daily outputs are expressed as anomalies from historical normals that were computed using 20 years (1982-2001) of satellite and surface weather data. TOPS is also capable of using short/mid-term weather/climate forecasts to produce forecasts of land surface conditions (snow pack, runoff, soil moisture and primary production) that are useful in resource management.

The first Moderate Resolution Imaging Spectroradiometer (MODIS) instrument is planned for launch by NASA in 1998. This instrument will provide a new and improved capability for terrestrial satellite remote sensing aimed at meeting the needs of global change research. The MODIS standard products will provide new and improved tools for moderate resolution land surface monitoring. These higher order data products have been designed to remove the burden of certain common types of data processing from the user community and meet the more general needs of global-to-regional monitoring, modeling, and assessment. The near-daily coverage of moderate resolution data from MODIS, coupled with the planned increase in high-resolution sampling from Landsat 7, will provide a powerful combination of observations. The full potential of MODIS will be realized once a stable and well-calibrated time-series of multispectral data has been established. In this paper the proposed MODIS standard products for land applications are described along with the current plans for data quality assessment and product validation.

A simple method for the estimation of global leaf area index (LAI) and fraction of photosynthetically active radiation absorbed by the vegetation (FAPAR) from atmospherically corrected Normalized Difference Vegetation Index (NDVI) observations is described. Recent improvements to the authors' three dimensional radiative transfer model of a vegetated surface are described. Example simulation results and a validation exercise are discussed. The model was utilized to derive land cover specific NDVI-LAI and NDVI-FAPAR relations. The method therefore requires stratification of global vegetation into cover types that are compatible with the radiative transfer model. Such a classification based on vegetation structure is proposed and a simple method for its derivation is presented. Proof-of-concept results are given to illustrate the feasibility of the proposed method.