Jump to Main Content
An assessment of the differences between spatial resolution and grid size for the SMAP enhanced soil moisture product over homogeneous sites
- A. Colliander, T.J. Jackson, S.K. Chan, P. O'Neill, R. Bindlish, M.H. Cosh, T. Caldwell, J.P. Walker, A. Berg, H. McNairn, M. Thibeault, J. Martínez-Fernández, K.H. Jensen, J. Asanuma, M.S. Seyfried, D.D. Bosch, P.J. Starks, C. Holifield Collins, J.H. Prueger, Z. Su, E. Lopez-Baeza, S.H. Yueh
- Remote sensing of environment 2018 v.207 pp. 65-70
- data collection, remote sensing, satellites, soil water
- Satellite-based passive microwave remote sensing typically involves a scanning antenna that makes measurements at irregularly spaced locations. These locations can change on a day to day basis. Soil moisture products derived from satellite-based passive microwave remote sensing are usually resampled to a fixed Earth grid that facilitates their use in applications. In many cases the grid size is finer than the actual spatial resolution of the observation, and often this difference is not well understood by the user. Here, this issue was examined for the Soil Moisture Active Passive (SMAP) enhanced version of the passive-based soil moisture product, which has a grid size of 9-km and a nominal spatial resolution of 33-km. In situ observations from core validation sites were used to compute comparison metrics. For sites that satisfied the established reliability and scaling criteria, the impact of validating the 9-km grid product with in situ data collected over a 9-km versus a 33-km domain was very small for the sites studied (0.039 m³/m³ unbiased root mean square difference for the 9-km case versus 0.037 m³/m³ for the 33-km case). This result does not mean that the resolution of the product is 9-km but that for the conditions studied here the soil moisture estimated from in situ observations over 9-km is a close approximation of the soil moisture estimated from in situ observations over the 33-km resolution. The implication is that using the enhanced SMAP product at its grid resolution of 9-km should not introduce large errors in most applications.