Examining the potential and effectiveness of water indices using multispectral sentinel-2 data to detect soil moisture as an indicator of mudflow occurrence in arid regions

• Published in: International journal of applied earth observation and geoinformation Vol. 130; p. 103887
• Main Authors: Al-Ali, Zahraa, Abulibdeh, Ammar, Al-Awadhi, Talal, Mohan, Midhun, Al Nasiri, Noura, Al-Barwani, Mohammed, Al Nabbi, Sara, Abdullah, Meshal
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.06.2024; Elsevier
• Subjects: Flood; Shaheen cyclone; Soil-line method; Spectral indices; Spectral reflectance curve; Spectral reflectance curve; Spectral indices; Shaheen cyclone; Soil-line method; Flood
• ISSN: 1569-8432
• DOI: 10.1016/j.jag.2024.103887

[Display omitted] •Multispectral Sentinel-2 data provided adequate spatial and temporal resolution to predict and monitor soil moisture.•Spectral water indices can be utilized as soil moisture indicators to predict high-risk mudflow phenomena in arid regions.•Indirect techniques can potentially validate the performance of spectral indices in inaccessible and large-scale areas. This study aims to evaluate the performance and effectiveness of six spectral water indices - derived from Multispectral sentinel-2 data - to detect soil moisture and inundated area in arid regions to be used as an indicator of mudflow phenomena to predict high-risk areas. Herein, the validation of the performance of spectral indices was conducted using threshold method, spectral curve performance, and soil-line method. These indirect validation techniques play a key role in saving time, effort, and cost, particularly for large-scale and inaccessible areas. It was observed that the Normalized Difference Water Index (NDWI), Modified Normalized Difference Water Index (mNDWI), and RSWIR indices have the potential to detect soil moisture and inundated areas in arid regions. According to the temporal spectral curve performance, the spectral characteristics of water and soil moisture were distinct in the Near infrared (NIR) and Short-wave Infrared (SWIR1,2) bands. However, the rate and degree differed between these bands, depending on the amount of water in the soil. Furthermore, the soil line method supported the appropriate selection of threshold values to detect soil moisture. However, the threshold values varied with location, time, season, and between indices. We concluded that considering the factors influencing the behavior of water and soil reflectivity could support decision-makers in identifying high-risk mudflow locations.