Date of Award

Spring 1-1-2017

Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

Eric E. Small

Second Advisor

Ethan Gutmann

Third Advisor

Ben Livneh

Fourth Advisor

Shemin Ge

Fifth Advisor

Balaji Rajagopalan


Soil moisture content plays a central role in the coupled water and energy exchange between the land surface and the atmosphere. It also controls infiltration rates and is therefore key to predicting groundwater recharge and discharge. Land Surface Models (LSMs) use meteorologic data with parameterizations of local soil and vegetation conditions to simulate soil moisture, runoff, and turbulent fluxes. Accurate predictions of droughts, floods, crop productivity, and climate change depend on our ability to understand and model the state and dynamics of surface soil moisture.

Satellite-based remote sensing missions provides global coverage and therefore offer the potential to improve existing LSMs. We use remotely-sensed and in situ soil moisture observations from seven well-instrumented field sites to estimate soil hydraulic properties (SHPs) in the Noah LSM. Default SHPs are based on mapped soil type, but ample evidence shows that soil type is a poor predictor of hydraulic behavior. Improvements can be made by calibrating these parameters to unbiased observations of surface soil moisture, especially when the dynamics of the default model are poor.

Remotely-sensed soil moisture observations measure between the surface and up to 5 cm depth. However, the shallowest layer of most LSMs and the placement of in situ probes is typically centered at 5 cm. This depth discrepancy affects observations of soil moisture dynamics. We find that after rain events, NASA’s SMAP (Soil Moisture Active Passive) satellite observes drying to occur over a 44% shorter timescale and twice as fast as 17 in situ validation networks spread across the globe.

Lastly, we demonstrate the strengths of SMAP and document how it differs from Noah simulated soil moisture over North America during drydown periods. Both SMAP and Noah drying rates depend on potential evaporation, soil moisture content, and vegetation. SMAP retrievals show that areas with sparse vegetation dry faster than areas with dense vegetation. Noah simulations show the opposite. After normalizing by potential evaporation, however, both SMAP and Noah data show that increased vegetation cover corresponds with lower evaporative efficiency. These differences are related to sensing depth and may also provide indications for how models can be improved.