SMOPS - Algorithm Description
SMOPS was developed to serve as a one-stop shop for all available operational satellite soil moisture products from individual sensors. It either ingests Level 2 soil moisture products from individual sensors for direct inclusion or retrieves soil moisture from L1B brightness temperature data from available low-frequency microwave satellite sensors using the Single Channel Retrieval (SCR) algorithm (Jackson, 1993).
SMOPS also offers a blended soil moisture data layer by merging soil moisture retrievals from all individual sensors using Cumulative Distribution Function (CDF) matching method at pixel level. Compared to soil moisture products from individual sensors, SMOPS blended data provides significantly better spatial coverage and, therefore, higher data availability for global applications.
The system has evolved through several versions to maintain a high-quality baseline and improve data timeliness. In Version 1.0, the baseline was switched to WindSat on the Coriolis satellite following the AMSR-E failure. Version 2.0 transitioned to AMSR-2 on GCOM-W1 as the baseline sensor. Additionally, the SCR algorithm is applied to near-real-time SMOS data to improve latency. In version 3.0, the SCR algorithm was expanded to include SMAP, real-time SMAP, and GMI on GPM to further enhance coverage and latency. In version 4.0, SMOS was dropped because of its designed lifespan and ASCAT-C was added as a new input sensor.
SMOPS generates two sets of global data products: a Daily Gridded Product with latency of 5 hours and a 6-Hour Gridded Product with latency of 3 hours. Each includes a blended product merging all available retrievals at each grid surface, alongside retrievals from all individual sensors. The Daily product covers the past 24 hours, while the 6-Hour product covers the past 6 hours.
Further details regarding the SCR algorithm, its sensitivity, error budgets, and merging methods are available in the Algorithm Theoretical Basis Document.