Close Window

CD-36 Abstract

Analyzing and Synthesizing LBA Field Data to Improve the Performance of Land Surface Models

Luis A Bastidas — Utah State University (US-PI)
Carlos Afonso Nobre — INPE - Instituto Nacional de Pesquisas Espaciais (SA-PI)
William James Shuttleworth — University of Arizona (US-PI)

This proposal is a direct extension of a proposal previously funded by NASA Ecology and Land Surface Hydrology Programs and will extend and further develop the data analysis and synthesis activities of that project. The focus is on making best use of the data gathered at the LBA field sites to improve the Land Surface Models (LSMs) ability to represent the surface-atmosphere interactions of Amazonian land covers in climate and weather prediction models and Land Data Assimilations Systems (LDAS). In our previous project, we developed and demonstrated a successful application of multi-criteria (MC) theory to the evaluation of LSMs using selected LBA data and, in particular, we demonstrated that MC theory, in the form of reliable computer algorithms, when applied to LBA data are able to: (a) identify parameters to which the performance of the LSM is sensitive and therefore merit optimization relative to data; (b) define realistic sets of model parameters whose use improves (typically by a factor of two, or greater) the LSMs simulations; (c) identify inadequacies in the formulation of LSMs that inhibit their ability to represent surface exchanges even with optimized parameters; and (d) investigate the extent to which LSM parameters differ between sites and/or before and after forest manipulations (e.g. selective logging) and, in this way, investigate differences in canopy function that might otherwise be confused by differences in the ambient conditions to which the canopies are exposed. This proposal will extend the application of the developed methods to all the quality-controlled data available from LBA (1) to improve the performance of selected LSMs, including SiB2 and SSiB, widely used in Brazil, and CLM and NOAH (with a carbon model added), widely used in regional and climate models; (2) to investigate site-to-site differences in parameters, the impact of data error on the parameters, and the value of remote sensing data in understanding these differences; and (3) in collaboration with a Brazilian co-investigator based in the NASA GLDAS group, to facilitate the development and improvement of a South American LDAS that can be used to document in near real time the continent-wide soil moisture fields and surface exchanges, including carbon exchange. By analyzing and synthesizing LBA data to evaluate and improve LSMs, this proposal will support the LBA programís efforts to address questions related to the priority topics Amazonian carbon dynamics, whole system functioning, and hydrometeorology and, by comparing different LSMs with different datasets, it will also provide an independent check on the reliability and value of relevant LBA data sets. The proposed project will continue the successful program of on-the-job training of Brazilian students, exchange visits between US and Brazilian researchers, and participation by leading PIs in the overall management of LBA through involvement in the LBA-SSC that were undertaken in our previous project.

Close Window