Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #312678

Research Project: Leveraging Remote Sensing, Land Surface Modeling and Ground-based Observations ... Variables within Heterogeneous Agricultural Landscapes

Location: Hydrology and Remote Sensing Laboratory

Title: Benchmarking the performance of a land data assimilation system for agricultural drought monitoring

Author
item Crow, Wade
item HAN, E. - Columbia University

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 12/15/2014
Publication Date: 1/1/2014
Citation: Crow, W.T., Han, E. 2015. Benchmarking the Performance of a Land Data Assimilation System for Agricultural Drought Monitoring [abstract]. American Geophysical Union, Fall Meeting Supplement. Abstract H12E-03.

Interpretive Summary:

Technical Abstract: The application of land data assimilation systems to operational agricultural drought monitoring requires the development of (at least) three separate system sub-components: 1) a retrieval model to invert satellite-derived observations into soil moisture estimates, 2) a prognostic soil water balance model to serve as the assimilation model, and 3) a sequential filter to update the assimilation model using the soil moisture retrievals. Each of these three components, in turn, requires a myriad of (often arbitrary) choices regarding their appropriate level of complexity and subsequent parameterization. Using a benchmarking strategy, this presentation will attempt to evaluate the performance of each of these sub-components relative to a baseline approach derived via simple linear regression of each sub-component’s inputs. Evaluation will be based on assessing the degree to which individual root-zone soil moisture predictions (or proxies thereof) can adequately anticipate temporal anomalies in the Normalized Difference Vegetation Index. Variations in benchmarking performance for various sub-components can then be attributed to the success of their non-linear parameterization (versus a simple linear baseline). Results will highlight the loss of skill associated with the imposition of non-linear physics into a soil water balance model and offer evidence that large-scale land data assimilation system are commonly poorly parameterized. Once indentified, these weaknesses can be systematically addressed in the development of next-generation drought monitoring systems.