Mapping man-made CO2 emissions using satellite-observed nighttime lights

Abstract:

The Open-Data Inventory for Anthropogenic Carbon dioxide (ODIAC) is a global high spatial resolution (1x1km) emission dataset for CO₂ emissions from fossil fuel combustion. The original version of ODIAC was developed at the Japanese Greenhouse Gas Observing Satellite (GOSAT) project to prescribe their inverse model. ODIAC first introduced the combined use of satellite-observed nighttime light data and individual power plant emission/geolocation information to estimate the spatial extent of fossil fuel CO₂. The ODIAC emission data has been widely used by the international carbon cycle research community and appeared in a number of publications in the literature. Since its original publication in 2011, we have made numerous modifications to the ODIAC emission model and the emission data have been updated on annual basis. We are switching from BP statistical data based emission estimates to estimates made by Carbon Dioxide Information Analysis Center (CDIAC) at Oak Ridge National Laboratory. In recent versions of ODIAC data, the emission seasonality has been adopted from the CDIAC monthly emission dataset. The emissions from international bunkers, which are not included in the CDIAC gridded emission data, are estimated using the UN Energy Database and included with the spatial distributions. In the next version of ODIAC emission model, we will explore the use of satellite data collected by the NASA’s Suomi National Polar-orbiting Partnership (NPP) satellite. We will estimate emission spatial distributions using global 500x500m nighttime lights data created from VIIRS data. We will also utilize a combustion detection algorithm Nightfire developed at NOAA National Geophysical Data Center to map gas flaring emissions. We also plan to expand our two emission sector emission distributing approach (power plant emission and non-point source emissions) by introducing a transportation emission sector which should improve emission distributions in urban and rural areas.