Publications in Scientific Journals:

F. Greifeneder, E. Khamala, D. Sendabo, W. Wagner, M. Zebisch, H. Farah, C. Notarnicola:
"Detection of soil moisture anomalies based on Sentinel-1";
Physics and Chemistry of the Earth, 112 (2019), 75 - 82.

English abstract:
Active and passive microwave remote sensing has evolved into a well-accepted and widely used method for the spatially continuous, coarse to medium resolution mapping of the surface soil moisture content (SMC). Presently, the exploitation of high-resolution data is less mature. Sentinel-1 is a high-resolution Synthetic Aperture Radar (∼5 by 20 m), which acts as the basis for mapping of SMC in this study. There aren't any other comparable SAR missions currently available whose data can be accessed and used for free.

One of the applications for SMC measurements is connected to the relationship between SMC anomalies and natural hazards such as droughts, flooding, or land-slides. A requirement for the detection and quantification of an anomaly is a long time-series (often 10-30 years) to derive a reference value. Herein lies one of the issues of Sentinel-1 based SMC mapping - at the time of writing, the Sentinel-1 time-series spanned a period of approximately 3.5 years.

We introduce an approach to overcome this problem and enable the Sentinel-1 based SMC anomaly detection. The method is based on a cross-calibration between Sentinel-1 SMC estimations and coarse resolution (∼30 km) modelled SMC from the Global Land Data Assimilation System (GLDAS), which covers the time-span from 1948 to today. As a result, we can derive the long-term averages for each Sentinel-1 pixel.

Results show that the proposed method allows a very accurate reproduction of the average SMC (for any given pixel) - if computed for the S1 time-span, the RMSE between estimated (GLDAS based) and true average S1 SMC is 0.7 %-Vol. Furthermore, the comparison with an in-situ time-series shows the correct detection of negative and positive anomalies, respectively. The method presented here may allow the integration of S1 data into, e.g., drought monitoring or flood forecasting applications.

"Official" electronic version of the publication (accessed through its Digital Object Identifier - DOI)

Created from the Publication Database of the Vienna University of Technology.