The Value of Data Assimilation for Dynamic Vegetation Monitoring in Land Surface Modeling

Abstract:

Globally, transpiration accounts for more than four-fifths of total evaporative flux (Jasechko et al., 2013), however virtually none of the major terrestrial hydrology land data assimilation systems (LDAS) explicitly simulate vegetation dynamics. This precludes assimilation of any of the plethora of the high-quality vegetation-monitoring products that are available from various remote-sensing platforms. In majority, existing LDAS rely on prescribed vegetation indexes, and thus cannot adapt either to landcover changes, or to dynamic vegetation response to terrestrial hydrology.

Recently, the Noah land surface model was extended into a multi-physics platform (Noah-MP) that includes a dynamic vegetation component (Niu et al., 2011). This model now has the potential to facilitate assimilation of remote sensing vegetation products into terrestrial hydrologic forecast and monitoring systems – especially ones that currently utilize the Noah LSM (e.g., Ek et al., 2003, Xia et al., 2011, Case et al., 2011).

In this talk, we explore the assimilation of vegetation-related information into Noah-MP. We start by assimilating in-situ observations of sensible and latent heat fluxes to update soil water and plant carbon states, and investigate how this type of relatively precise data assimilation can be useful for exposing deficiencies in the new dynamic vegetation model. We then assimilate EVI observations from MODIS (MOD13 and MOD15), and explore the potential for satellite-observed changes in vegetation states to impact hydrological predictions. All model estimates are validated against FluxNet tower data.