Wetland Species Mapping Using Advanced Technological Measurement

Abstract

Wetlands are pivotal in supporting the natural ecosystem and maintaining biodiversity while being susceptible to anthropogenic activities and climate change. However, monitoring wetlands over a large geographical and temporal extent is challenging. Vegetation health can be considered a good indicator of wetland conditions, and measuring chlorophyll content will provide insight into vegetation health. Linking wetland species mapping from chlorophyll spectral indices to local and regional conservation strategies could improve biodiversity conservation. Here, we apply this to Keetham Lake, India, using machine learning methods (relevance vector model) and hyperspectral measurements. From 10 chlorophyll‐sensitive spectral indices, we identified four as best performing, particularly for: TVI + CCCI + NDRE for calibration and NDRE + TVI for validation data. The least performing combinations were MCARI for calibration and TVI + CCCI + NDRE + MCARI for validation. Overall, we identified that NDRE + TVI was the best‐performing pair of spectral indices for chlorophyll assessment and implementation in wetland species. This approach allows for precise mapping of wetland species, providing data on their extent and the area they cover. By creating a digital database, this method enables long‐term monitoring of changes in wetland species' numbers and distribution, helping to assess trends of increase or decline in freshwater ecosystems. Such strategies are vital for supporting both local and global conservation efforts, offering insights for forward‐looking, data‐driven preservation initiatives.