Yudong Tian

Land information system: An interoperable framework for high resolution land surface modeling

Abstract Knowledge of land surface water, energy, and carbon conditions are of critical importance due to their impact on many real world applications such as agricultural production, water resource management, and flood, weather, and climate prediction. Land Information System (LIS) is a software framework that integrates the use of satellite and ground-based observational data along with advanced land surface models and computing tools to accurately characterize land surface states and fluxes. LIS employs the use of scalable, high performance computing and data management technologies to deal with the computational challenges of high resolution land surface modeling. To make the LIS products transparently available to the end users, LIS includes a number of highly interactive visualization components as well. The LIS components are designed using object-oriented principles, with flexible, adaptable interfaces and modular structures for rapid prototyping and development. In addition, the interoperable features in LIS enable the definition, intercomparison, and validation of land surface modeling standards and the reuse of a high quality land surface modeling and computing system.

Component analysis of errors in satellite-based precipitation estimates

[1]xa0Satellite-based precipitation estimates have great potential for a wide range of critical applications, but their error characteristics need to be examined and understood. In this study, six (6) high-resolution, satellite-based precipitation data sets are evaluated over the contiguous United States against a gauge-based product. An error decomposition scheme is devised to separate the errors into three independent components, hit bias, missed precipitation, and false precipitation, to better track the error sources associated with the satellite retrieval processes. Our analysis reveals the following. (1) The three components for each product are all substantial, with large spatial and temporal variations. (2) The amplitude of individual components sometimes is larger than that of the total errors. In such cases, the smaller total errors are resulting from the three components canceling one another. (3) All the products detected strong precipitation (>40 mm/d) well, but with various biases. They tend to overestimate in summer and underestimate in winter, by as much as 50% in either season, and they all miss a significant amount of light precipitation (<10 mm/d), up to 40%. (4) Hit bias and missed precipitation are the two leading error sources. In summer, positive hit bias, up to 50%, dominates the total errors for most products. (5) In winter, missed precipitation over mountainous regions and the northeast, presumably snowfall, poses a common challenge to all the data sets. On the basis of the findings, we recommend that future efforts focus on reducing hit bias, adding snowfall retrievals, and improving methods for combining gauge and satellite data. Strategies for future studies to establish better links between the errors in the end products and the upstream data sources are also proposed.

High-performance Earth system modeling with NASA/GSFC’s Land Information System

The Land Information System software (LIS; http://lis.gsfc.nasa.gov/, 2006) has been developed to support high-performance land surface modeling and data assimilation. LIS integrates parallel and distributed computing technologies with modern land surface modeling capabilities, and establishes a framework for easy interchange of subcomponents, such as land surface physics, input/output conventions, and data assimilation routines. The software includes multiple land surface models that can be run as a multi-model ensemble on global or regional domains with horizontal resolutions ranging from 2.5° to 1xa0km. The software may execute serially or in parallel on various high-performance computing platforms. In addition, the software has well-defined, standard-conforming interfaces and data structures to interface and interoperate with other Earth system models. Developed with the support of an Earth science technology office (ESTO) computational technologies project round~3 cooperative agreement, LIS has helped advance NASA’s Earth–Sun division’s software engineering principles and practices, while promoting portability, interoperability, and scalability for Earth system modeling. LIS was selected as a co-winner of NASA’s 2005 software of the year award.

Systematic anomalies over inland water bodies in satellite‐based precipitation estimates

[1]xa0We studied two recent high-resolution precipitation datasets derived from satellite remote-sensing, TRMM 3B42 and CMORPH, and compared them with ground-based radar and gauge measurements in the southeastern U. S. We found there are systematic differences in rainfall estimates between inland water-body pixels and land pixels. On average, there are about twice as many raining days over water bodies than over land pixels in the satellite products, causing much higher false alarm rates over water bodies. The increased false alarms occur mostly in the form of light rain (<2 mm/day), and lead to significantly different rain rate distributions between water-body and land pixels. We speculate that this inconsistency is caused by deficiencies in emissivity characterization for the passive microwave-based rainfall retrievals that serve as input to these merged products.

Evaluation of a satellite-based global flood monitoring system

This study provides an initial evaluation of a global flood monitoring system (GFMS) using satellite-based precipitation and readily available geospatial datasets. The GFMS developed by our group uses a relatively simple hydrologic model, based on the run-off curve number method, to transform precipitation into run-off. A grid-to-grid routing scheme moves run-off downstream. Precipitation estimates are from the TRMM Multi-satellite Precipitation Analysis (TMPA). We first evaluated the TMPA algorithm using a radar/gauge merged precipitation product (Stage IV) over south-east USA. This analysis indicated that the spatial scale (and hence the basin size) as well as regional and seasonal considerations are important in using the TMPA to drive hydrologic models. GFMS-based run-off simulations were evaluated using observed streamflow data at the outlet of two US basins and also using a global flood archive. Basin-scale analysis showed that the GFMS was able to simulate the onset of flood events produced by heavy precipitation; however, the simulation performance deteriorated in the later stages. This result points out the need for an improved routing component. Global-scale analysis indicated that the GFMS is able to detect 38% of the observed floods; however, it suffers from region-dependent bias.

High-performance land surface modeling with a Linux cluster

The Land Information System (LIS) was developed at NASA to perform global land surface simulations at a resolution of 1-km or finer in real time. Such unprecedented scales and intensity pose many computational challenges. In this article, we demonstrate some of our approaches in high-performance computing with a Linux cluster to meet these challenges and reach our performance goals. These approaches include job partition and a job management system for parallel processing on the cluster, high-performance parallel input/output based on GrADS-DODS (GDS) servers, dynamic load-balancing and distributed data storage techniques, and highly scalable data replication with peer-to-peer (P2P) technology. These techniques work coherently to provide a high-performance land surface modeling system featuring fault tolerance, optimal resource utilization, and high scalability. Examples are given with LISs high-resolution modeling of surface runoff during 2003 to illustrate LISs capability to enable new scientific explorations.

A prototype of land information sensor web (LISW)

To meet future earth system science challenges, NASA will develop constellations of smart satellites in sensor web configurations that provide timely on-demand data and analysis to users, and that be reconfigured based on the changing needs of science and available technology. Sensor webs can eclipse the value of disparate sensor components by reducing response time and increasing scientific value, especially when integrated with science analysis, data assimilation, prediction modeling and decision support tools. The prototype Land Information Sensor Web (LISW) is a project sponsored by NASA, trying to integrate the Land Information System (LIS) in a sensor web framework which allows for optimal 2-way information flow that enhances land surface modeling using sensor web observations, and in turn allows sensor web reconfiguration to minimize overall system uncertainty. This prototype is based on a simulated interactive sensor web, which is then used to exercise and optimize the sensor web - modeling interfaces. These synthetic experiments provide a controlled environment in which to examine the end-to-end performance of the prototype, the impact of various design sensor web design trade-offs and the eventual value of sensor webs for particular prediction or decision support. The Study of virtual Land Information Sensor Web (LISW) is expected to provide some necessary priori knowledge for designing and deploying the next generation Global Earth Observing System of systems (GEOSS). In this paper, the progress of the LISW study will be presented, especially in scenario experiment design, sensor web framework and uncertainties in current land surface modeling.