Menu
Increased availability of and access to satellite and remotely sensed data have paved the way for innovations in methods and tools for climate evaluation. These have allowed us to observe, measure, and evaluate the impact of climate change on humans and ecosystems at a relatively low cost. High-resolution and frequent observations from satellites have facilitated the creation of wide ranging and precise proxy indicators (complementary to traditional on-the-ground measures) to observe environmental outcomes such as carbon emissions, land use changes, and vegetation health. The volume of available data has enabled the training of robust machine learning models, allowing for more accurate classification, detection, and prediction tasks in geospatial analysis.
In this presentation, we explore the pivotal role of these innovations in enhancing climate evaluation methodologies and driving informed, evidence-based decision-making in climate policy. In doing so, we share an interface called “Remote Sensing Indicators for Development”, a resource compiling existing literature, empirical applications, and guidance on the use of remote sensing and machine learning in evidence generation and synthesis. We showcase the multifaceted ways in which researchers may use this resource in their own work – to further their understanding of the distinctions and the comparability between various remotely sensed indices, including their advantages and disadvantages, and to determine which indices may be used for ML-based prediction in regions where data is lacking. This presentation fills a critical knowledge gap by offering a comprehensive resource that explores the spectrum of applications that leverage remote sensing indicators and machine learning in climate evaluation.