Performance of CCCma and GFDL climate models using remote sensing and surface data for the state of Rio de Janeiro-Brazil

The performance of the CCCma (Canadian Centre for Climate Modelling and Analysis) and GFDL (Geophysical Fluid Dynamic Laboratory) models in the baseline period (1961–2000) and for the future IPCC scenario Representative Concentration Pathways (RCP) 6.0 (2046–2065) were evaluated through the descriptive statistics using spatial interpolation methods (Kriging and Ordinary Co-Kriging) using the exponential, gaussian and spherical spatial models. The daily temperature and rain data from the CCCma and GFDL models were used for the current baseline climate (1961–2000) and the future scenario (2046–2065), data from the National Oceanic and Atmospheric Administration (NOAA) were used for comparisons with the two models and the product MCD12Q1 derived from Moderate resolution spectroradiometer sensor was used to check which areas had higher and lower values of temperature and rain in the State of Rio de Janeiro. For validation, data were compared and geostatistical analysis was performed using Kriging techniques. The averages of air temperature and precipitation showed similar patterns in both models. The CCCma model presented the data closest to the NOAA reanalysis data, with the GFDL model underestimating most precipitation and air temperature data. Through the product MCD12Q1 of the MODIS sensor, it is possible to register marked differences in relation to rain and temperature and their respective land use for the State. For the geostatistics data, it was possible to verify that, for the past and future scenarios, the exponential transitive model presented in the majority of the cases the least degree of spatial dependence (GDE), being therefore considered the best. When comparing the GDE of the two climatic models, it is verified that the CCCma presented the best geostatistical performance and can be used in works to simulate scenarios of future climate changes. [Display omitted] •The use of climate models associated with remote sensing is important.•Space models are important for understanding climate change.•Spatial models represented the physiography of the state of Rio de Janeiro well.•The use of MCD12Q1 and climate models may serve to monitor cities in the future.