Fusion of SAR and optical remote sensing data to map montane forest biomass: A case study in El Choco, Ecuador

Optical remote sensing, which includes visible, near-infrared and shortwave-infrared, has been widely used to estimate aboveground biomass (AGB) of different types of forests in diverse environmental conditions. AGB refers to all vegetation above the ground, including stems, branches, bark, seed, and foliage of live plants. This measure also provides an estimate of forest carbon storage capacity (50% of biomass according to IPCC standard value). However, in cloudy areas, the availability of free-cloud satellite data can become limited or null, depending on seasonality and local weather conditions. In Ecuador, the tropical Andes region, comprises a geographical extension from the Andean highlands to the Coast, and annual cloud cover can averages around 90%. Synthetic Aperture Radar (SAR) systems ignore cloud coverage and provide estimates of forest height. In this study, we proposed the fusion of active remote sensing data, specifically the Copernicus Sentinel-1 SAR C-band Ground Range Detected (GRD), with optical data from Sentinel-2A+B Multispectral Instrument (MSI) to estimate AGB of Choco montane forests, located in the tropical North Western region of Ecuador. We generated seasonal and intra-annual time series of AGB estimates for 2018, and finally we compared our results with studies that used exclusively passive remote sensing data, and in-situ data. The results indicated the advantages of using SAR fused with optical remote sensing data for AGB estimation, especially in wet season where cloud coverage can be permanent.