Assessing space-based smoldering peatland in the tropics with atmospheric products from multi-sensor satellites

• Published in: Modeling earth systems and environment Vol. 10; no. 1; pp. 465 - 481
• Main Authors: Sofan, Parwati, Chulafak, Galdita A., Yulianto, Fajar, Suwarsono, Suwarsono, Tonoto, Prayoto, Prasasti, Indah, Khomarudin, M. Rokhis, Adiningsih, Erna Sri
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.02.2024; Springer Nature B.V
• Subjects: Aerosols; Algorithms; Atmospheric data; Carbon monoxide; Chemistry and Earth Sciences; Computer Science; Confidence intervals; Disaster management; Earth and Environmental Science; Earth Sciences; Earth System Sciences; Ecosystems; Emergency preparedness; Environment; Landsat; Low temperature; Math. Appl. in Environmental Science; Mathematical Applications in the Physical Sciences; Monitoring instruments; Optical analysis; Optical thickness; Original Article; Peatlands; Physics; Pixels; Remote sensing; Satellite imagery; Satellites; Sensors; Smoldering; Spatial discrimination; Spatial resolution; Spectroradiometers; Statistical analysis; Statistics for Engineering; Tropical environments; Wind speed; Landsat-8; Smoldering tropical peatland; TROPOMI; MODIS; SGLI; Indonesia
• DOI: 10.1007/s40808-023-01793-4
• ISSN: 2363-6203

The research on space-based tropical smouldering peatlands poses challenges due to their low temperature and background object environments. This study focuses on analyzing smoldering pixels detected by the Tropical Peatland Combustion Algorithm (ToPeCAl) applied to Landsat-8 imagery in Indonesian peatlands. A modification to the contextual test for selecting fire pixels is proposed. Instead of using the high confidence level typically employed for selecting flaming pixels, a low confidence level is suggested to identify smoldering pixels from their background pixels. The validation of smoldering pixels is conducted using atmospheric products from multi-sensor satellites. The findings reveal that applying the low confidence level on the contextual test results in an accuracy of 75% for smoldering detection, similar to previous research with ground check validation. A dynamic statistical threshold value is utilized from background pixels within a moving window to validate smoldering pixels using atmospheric data, taking into consideration factors such as wind velocity, spatial resolution, and time of acquisition between sensors. Approximately 50–70% of the selected smoldering pixels show higher values of Aerosol Optical Depth from Moderate Resolution Imaging Spectroradiometer (MODIS-AOD), Carbon Monoxide from TROPOspheric Monitoring Instrument (TROPOMI-CO), and Absorbing Aerosol Index from Second-generation Global Imager (SGLI-AAI) compared to their mean background pixel values. The study reveals a significant correlation (up to 0.8) between TROPOMI-CO and ground measurement of AOD over smouldering peatlands. These insights have the potential to aid disaster management agencies in effectively controlling smoldering fires and minigating the risk associated with smoldering peatlands.