Scaling and Assimilation of Soil Moisture and Streamflow

Christoph Ruediger, Jeffrey Walker, Garry Willgoose, Jetse Kalma, Riki Davidson, Manju Hemakumara, Paul Houser and Ross Woods

Power Point Presentation

The aim of this project is to provide the best estimate of spatial distribution and temporal variation of soil moisture content in the root zone through a combination of modelling and observations using data assimilation. Satellite passive microwave remote sensing (specifically AMSR-E) provides an estimate of the average near-surface soil moisture content over an area of approximately 25x25km2 for regions with low vegetation. The “scaling” component of SASMAS will develop techniques for downscaling the large-scale satellite measurements to the point-scale, while the “soil moisture assimilation” component will investigate the best way to use both the large- and point- scale near-surface soil moisture data to better estimate the root zone soil moisture content in areas of low vegetation. The “streamflow assimilation” component will investigate how streamflow data may be used to better estimate the root zone soil moisture content in areas of dense vegetation, and how it may be used to further constrain estimates in areas of low vegetation.

There is a large field effort in support of this project. The focus of this field work includes the validation of AMSR-E near-surface soil moisture data, collection of near-surface soil moisture and stream flow data for assimilation, and monitoring of point soil moisture profiles for validation purposes. The Goulburn River Catchment has been chosen for this study due to its proximity to Newcastle and high percentage of cleared agricultural land north of the Goulburn River, required for the AMSR-E remote sensing. Most of the research will be conducted in that region, with two smaller intensively monitored subcatchments located on the Krui and Merriwa Rivers, and even more intensive monitoring of a smaller subcatchment of the Krui River. The southern part of the catchment is mostly forested and will be monitored less intensively. There will be a total of 25 soil moisture profile monitoring sites, 6 stream gauges, 5 weather stations, and numerous rain gauges.

While field data is being collected, soil moisture and streamflow modelling will focus on using synthetic data sets, as well as existing data sets such as that for the Mahurangi catchment in New Zealand. The model will be a version of VIC-2L that includes a streamflow routing model. Streamflow assimilation will focus on using the Ensemble Kalman Filter and variational techniques, while the soil moisture assimilation will focus on using the Ensemble Kalman Filter. Current progress on both field work and modelling efforts will be reported.