Simulation and forecasting of atmospheric CO2 using RAMS

Concentration of carbon dioxide as the main greenhouse gas has been increasing, which has seri-ous impacts on climate change, ecosystem, and human health. This study uses the Regional At-mospheric Modeling System (RAMS V6.2) with the Simple Biosphere (SiB-2) sub-model to simu-late and forecast the concentration of carbon dioxide over the Daintree Rainforest in Australia. To evaluate the performance of the model, results are compared with the satellite product (MultiIn-strumentFusedXCO2). This work evaluates the impact of various grid sizes, different combina-tions of diffusion parameters, and boundary conditions on CO2 simulations. Results reveal that RAMS can remarkably improve the statistical measures of the simulations particularly compared with the previous modeling analysis. The maximum mean bias error and RMSE in this study for all simulations are smaller than 1.85 and 2.07 ppm, respectively. The most proper simulation type (type III) leads to the average bias error and RMSE of 0.34 and 0.94 ppm, respectively. Fur-thermore, results show that the fuzzy algorithm is a proper method to provide initial CO2 in fore-casting. This study shows that RAMS can be efficiently used in CO2 simulation and forecasting, which is valuable for CO2 monitoring, climate, environment, and management applications.