Estimating three-wheeler emissions in urban areas of the global south using remote sensing data, open data and GPS tracking data

Abstract

This research addresses the lack of data on 3-wheelers, their vehicle kilometers traveled (VKT), and resulting greenhouse gas (GHG) emissions. We outline a methodology using aerial imagery and deep learning to analyze vehicle fleet composition. Given the uncertainty around their numbers, we focus on 3-wheelers, demonstrating how GPS trajectory data can refine VKT estimates. By combining these findings with land use data and emission factors, we calculate the total CO2 emissions of 3-wheelers in a specific area. This research aims to improve our understanding of global vehicle fleets and their GHG impact. It is relevant not only for emission inventories but also for assessing climate change mitigation measures. This work is part of the Emissionslandkarte (ELK) project at the German Aerospace Center (DLR), which collects and estimates the impact of transport on GHG emissions globally (Righi et al., 2023).