Spatial distribution of XCO2 using OCO-2 data in growing seasons

Abstract

The purpose of this research is to assess the spatial distribution of CO₂ concentration during the growing seasons (April to September) in 2015 over Iran. The XCO₂ data belonging to orbiting carbon observatory-2 (OCO-2) and eight environmental variables data consist of normalized difference vegetation index (NDVI), net primary productivity (NPP), land surface temperature (LST), leaf area index (LAI), air temperature, wind speed, wind direction, and national land cover map were modeled by multi-layer perceptron (MLP). The values of R² and RMSE indices show the good performance of the multi-layer perceptron model for monthly models. Based on sensitivity analysis results, land cover and wind direction had the most important role in the spatial distribution of XCO₂. Also, the results revealed that the maximum values of XCO₂ observed in the east, south east, and desert areas in central of Iran due to the lack of vegetation cover, lack of local wind current, and high temperature. The western, northwestern and northern regions of Iran have the minimum amounts of XCO₂ because of existing valuable ecosystem such as Hyrcanian and Zagrous forests, rangeland, air currents, and low temperature. The findings of this study indicated that the manageable factors such as land cover and vegetation cover play very important roles in the spatial distribution of CO₂ and finding carbon dioxide source and sink at national scale. Therefore, policymakers and managers by the logical management of these resources are able to control or even reduce the concentration of carbon dioxide in different areas.

Keywords: Carbon dioxide; Iran; Land cover; Modeling.