Translating aboveground cosmic-ray neutron intensity to high-frequency soil moisture profiles at sub-kilometer scale

Authors

R. Rosolem, University of BristolFollow
T. Hoar, NCAR Data Assimilation Research Section
A. Arellano, University of Arizona
J. L. Anderson, NCAR Data Assimilation Research Section
W. J. Shuttleworth, University of Arizona
X. Zeng, University of Arizona
Trenton E. Franz, University of Nebraska–LincolnFollow

Document Type

Article

Date of this Version

11-4-2014

Citation

Author(s) 2014. CC Attribution 3.0 License

Comments

Hydrol. Earth Syst. Sci., 18, 4363–4379, 2014 www.hydrol-earth-syst-sci.net/18/4363/2014/ doi:10.5194/hess-18-4363-2014

Abstract

Above-ground cosmic-ray neutron measurements provide an opportunity to infer soil moisture at the subkilometer scale. Initial efforts to assimilate those measurements have shown promise. This study expands such analysis by investigating (1) how the information from aboveground cosmic-ray neutrons can constrain the soil moisture at distinct depths simulated by a land surface model, and (2) how changes in data availability (in terms of retrieval frequency) impact the dynamics of simulated soil moisture profiles. We employ ensemble data assimilation techniques in a “nearly-identical twin” experiment applied at semi-arid shrubland, rainfed agricultural field, and mixed forest biomes in the USA. The performance of the Noah land surface model is compared with and without assimilation of observations at hourly intervals, as well as every 2 days. Synthetic observations of aboveground cosmic-ray neutrons better constrain the soil moisture simulated by Noah in root-zone soil layers (0–100 cm), despite the limited measurement depth of the sensor (estimated to be 12–20 cm). The ability of Noah to reproduce a “true” soil moisture profile is remarkably good, regardless of the frequency of observations at the semi-arid site. However, soil moisture profiles are better constrained when assimilating synthetic cosmic-ray neutron observations hourly rather than every 2 days at the cropland and mixed forest sites. This indicates potential benefits for hydrometeorological modeling when soil moisture measurements are available at a relatively high frequency. Moreover, differences in summertime meteorological forcing between the semi-arid site and the other two sites may indicate a possible controlling factor to soil moisture dynamics in addition to differences in soil and vegetation properties.