On the identification of representative in situ soil moisture monitoring stations for the validation of SMAP soil moisture products in Australia

• Published in: Journal of hydrology (Amsterdam) 2016-06, Vol.537, p.367-381
• Main Authors: Yee, Mei Sun, Walker, Jeffrey P., Monerris, Alessandra, Rüdiger, Christoph, Jackson, Thomas J.
• Format: Article
• Language: English
• Subjects: agricultural land; Australia; geostatistics; Hydrology; Land; land use; Monitoring; radar; remote sensing; Representativeness; Satellite validation; Satellites; soil heterogeneity; Soil moisture; Soil Moisture Active Passive (SMAP); Soil moisture measurement; Soil stability; soil water; Spatial stability; Stability; Stations; statistical analysis; Temporal stability
• ISSN: 0022-1694; 1879-2707
• DOI: 10.1016/j.jhydrol.2016.03.060

•Representative stations according to different SMAP scales were identified.•Airborne soil moisture products are useful for identifying representative locations.•Recommendations were made for the design of future satellite validation networks. The high spatio-temporal variability of soil moisture complicates the validation of remotely sensed soil moisture products using in situ monitoring stations. Therefore, a standard methodology for selecting the most representative stations for the purpose of validating satellites and land surface models is essential. Based on temporal stability and geostatistical studies using long-term soil moisture records, intensive ground measurements and airborne soil moisture products, this study investigates the representativeness of soil moisture monitoring stations within the Yanco study area for the validation of NASA’s Soil Moisture Active Passive (SMAP) products at 3km for radar, 9km for radar–radiometer and 36km for radiometer pixels. This resulted in the identification of a number of representative stations according to the different scales. Although the temporal stability method was found to be suitable for identifying representative stations, stations based on the mean relative difference (MRD) were not necessarily the most representative of the areal average. Moreover, those identified from standard deviation of the relative difference (SDRD) may be dry-biased. It was also found that in the presence of heterogeneous land use, stations should be weighted based on proportions of agricultural land. Airborne soil moisture products were also shown to provide useful a priori information for identifying representative locations. Finally, recommendations are made regarding the design of future networks for satellite validation, and specifically the most representative stations for the Yanco area.