Skip to main content
ARS Home » Research » Publications at this Location » Publication #285504

Title: Benchmarking LSM root-zone soil mositure predictions using satellite-based vegetation indices

item Crow, Wade
item BOLTEN, J - National Aeronautics And Space Administration (NASA)
item KUMAR, S - Collaborator

Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 10/15/2012
Publication Date: 1/6/2013
Citation: Crow, W.T., Bolten, J., Kumar, S. 2013. Benchmarking LSM root-zone soil mositure predictions using satellite-based vegetation indices[abstrat]. 2011 SMOS Science Workshop. 2011 CDROM.

Interpretive Summary:

Technical Abstract: The application of modern land surface models (LSMs) to agricultural drought monitoring is based on the premise that anomalies in LSM root-zone soil moisture estimates can accurately anticipate the subsequent impact of drought on vegetation productivity and health. In addition, the water and energy balance functions of LSMs are widely assumed to add value to drought predictions. This, in turn, implies that LSM soil moisture outputs are more valuable for drought monitoring than simpler drought products based solely on the consideration of rainfall anomalies. With these assumptions in mind, this presentation benchmarks the performance of modern LSMs relative to a simple water accounting procedure based solely on observed precipitation (i.e., the well-known Antecedent Precipitation Index (API) model). In particular, the lagged rank cross-correlation between model-derived root-zone soil moisture estimates and remotely-sensed vegetation indices (VI) is examined between January 2000 and December 2010 to quantify the skill of modern LSMs (and a baseline API approach) for agricultural drought monitoring. Soil moisture products with the highest correlation versus future VI anomalies are assumed to contribute the most utility to an agricultural drought monitoring system. Results suggest that, when averaged in bulk across the annual cycle, little or no added skill (<5% in relative terms) is associated with applying modern LSMs to off-line agricultural drought monitoring relative to simple accounting procedures based solely on observed precipitation accumulations. However, slightly larger amounts of added skill (5-15% in relative terms) are identified when focusing exclusively on the extra-tropical growing season and/or utilizing root-zone soil moisture values acquired by averaging across a multi-model ensemble. Key presentation results will be verified via an independent analysis based on comparisons against satellite-derived surface soil moisture retrievals.