Effects of vegetation dynamics on evapotranspiration and soil moisture in northwestern Mexico

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 24 January 2011
Effects of vegetation dynamics on evapotranspiration and soil moisture in northwestern Mexico
Washington State Convention Center
Qiuhong Tang, University of Washington, Seattle, WA; and F. Munoz-Arriola, E. R. Vivoni, and D. P. Lettenmaier

Poster PDF (431.6 kB)

Evapotranspiration (ET) and soil moisture (SM) are closely related to vegetation dynamics in northwestern Mexico, a region characterized by an abrupt increase in rainfall and ecosystem green-up during the North American monsoon (NAM). Most land surface models use a fixed seasonal vegetation cycle and are therefore unable to fully capture the spatiotemporal changes in vegetation. In this study, time-varying leaf area index (LAI) and a fixed seasonal LAI cycle, both inferred from the Moderate Resolution Imaging Spectroradiometer (MODIS), were compared as inputs to the Variable Infiltration Capacity (VIC) model over northwestern Mexico during 2001-2008. Model results for the two sets of simulations were compared with latent heat fluxes observed by two eddy covariance tower sites for three summer periods. The results show that both vegetation greening onset and dormancy dates vary substantially from year to year with a range of more than half a month. Using the fixed season LAI cycle, the model tends to under- (over-) estimate ET and over- (under-) SM when vegetation greening occurs earlier (later) than the mean greening onset date. The discrepancies in SM were generally less than 2% when aggregated to 1 degree by 1 degree sub-domain and averaged over the summer season. In contrast, the discrepancies in ET were large, especially during a period of approximately two week at the beginning of the monsoon. The effect of vegetation dynamics on ET estimates was about 10% in the Sierra Madre Occidental and 30% in the continental interior east of the mountain range. VIC-estimated ET with the time-varying LAI had high interannual variability at the greening onset and dormancy periods corresponding to the vegetation dynamics. The greening onset date was highly related to ET early in the monsoon season, indicating the potential usefulness of LAI anomalies for predicting early season ET.