Evaluation of 18 satellite- and model-based soil moisture products using in situ measurements from 826 sensors

Abstract. Information about the spatiotemporal variability of soil moisture is critical for many purposes, including monitoring hydrologic extremes, irrigation scheduling, and prediction agricultural yields. We evaluated temporal dynamics 18 state-of-the-art (quasi-)global near-surface products, six based on satellite retrievals, models without data assimilation (referred to hereafter as “open-loop” models), that assimilate or brightness temperature data. Seven products are introduced first time in this study: one multi-sensor merged product called MeMo (Merged Moisture) estimates from HBV (Hydrologiska Byråns Vattenbalansavdelning) model with three precipitation inputs (ERA5, IMERG, MSWEP) SMAPL3E respectively. As reference, we used situ measurements between 2015 2019 at 5 cm depth 826 sensors, located primarily USA Europe. The 3-hourly Pearson correlation (R) was chosen primary performance metric. found application Soil Wetness Index (SWI) smoothing filter resulted improved all products. best-to-worst ranking four single-sensor SMAPL3ESWI, SMOSSWI, AMSR2SWI, ASCATSWI, L-band-based SMAPL3ESWI (median R 0.72) outperforming others 50 % sites. Among two (MeMo ESA-CCISWI), performed better average 0.72 versus 0.67), probably due inclusion SMAPL3ESWI. open-loop HBV-MSWEP, HBV-ERA5, ERA5-Land, HBV-IMERG, VIC-PGF, GLDAS-Noah. This largely reflects quality forcing. HBV-MSWEP 0.78) best not just among but calibration median by +0.12 compared random parameters, highlighting importance calibration. HBV-MSWEP+SMAPL3E, HBV-ERA5+SMAPL3E, GLEAM, SMAPL4, HBV-IMERG+SMAPL3E, ERA5. retrievals into HBV-IMERG +0.06, suggesting yields significant benefits global scale.