Soil moisture mapping using Sentinel 1 images: the proposed approach and its preliminary validation carried out in view of an operational product

S. Paloscia; S. Pettinato; E. Santi; N. Pierdicca; L. Pulvirenti; C. Notarnicola; G. Pace; A. Reppucci

doi:10.1117/12.899523

Translator Disclaimer

26 October 2011 Soil moisture mapping using Sentinel 1 images: the proposed approach and its preliminary validation carried out in view of an operational product

S. Paloscia, S. Pettinato, E. Santi, N. Pierdicca, L. Pulvirenti, C. Notarnicola, G. Pace, A. Reppucci

Author Affiliations +

Proceedings Volume 8179, SAR Image Analysis, Modeling, and Techniques XI; 817904 (2011) https://doi.org/10.1117/12.899523
Event: SPIE Remote Sensing, 2011, Prague, Czech Republic

Abstract

The main objective of this research is to develop, test and validate a soil moisture (SMC)) algorithm for the GMES Sentinel-1 characteristics, within the framework of an ESA project. The SMC product, to be generated from Sentinel-1 data, requires an algorithm able to process operationally in near-real-time and deliver the product to the GMES services within 3 hours from observations. Two different complementary approaches have been proposed: an Artificial Neural Network (ANN), which represented the best compromise between retrieval accuracy and processing time, thus allowing compliance with the timeliness requirements and a Bayesian Multi-temporal approach, allowing an increase of the retrieval accuracy, especially in case where little ancillary data are available, at the cost of computational efficiency, taking advantage of the frequent revisit time achieved by Sentinel-1. The algorithm was validated in several test areas in Italy, US and Australia, and finally in Spain with a 'blind' validation. The Multi-temporal Bayesian algorithm was validated in Central Italy. The validation results are in all cases very much in line with the requirements. However, the blind validation results were penalized by the availability of only VV polarization SAR images and MODIS lowresolution NDVI, although the RMS is slightly > 4%.

Citation Download Citation

S. Paloscia, S. Pettinato, E. Santi, N. Pierdicca, L. Pulvirenti, C. Notarnicola, G. Pace, and A. Reppucci "Soil moisture mapping using Sentinel 1 images: the proposed approach and its preliminary validation carried out in view of an operational product", Proc. SPIE 8179, SAR Image Analysis, Modeling, and Techniques XI, 817904 (26 October 2011); https://doi.org/10.1117/12.899523

ACCESS THE FULL ARTICLE