Pattern-based Calibrated Climate Forecasts in Guatemala: Implementation and Climate Services

The monthly forecast of rainfall and temperature is required by different sectors and users in Guatemala, specially by agriculture and disaster risk management. Considering that demand, the National Meteorological and Hydrological Service of Guatemala (INSIVUMEH), has implemented a seasonal and monthly forecast based on calibrated and operational global circulation models (GCMs) following the NextGen methodology, developed by Columbia University’s International Research Institute for Climate and Society (IRI). This work discusses the predictive skill of the pattern-based bias-corrected model ensembles utilized by INSIVUMEH, and also compares it with other methodologies such as climatology-based and analog years forecast, which have been previously used operationally in the country. For this evaluation, the rainfall and temperature output from the NMME and C3S models have been employed as predictors, while observations from the CHIRPS and locally-developed ENACTS datasets have been used for calibration in the canonical-correlation analysis (CCA) executed to generate the forecast. It is also discussed how this forecast has been implemented as a probabilistic forecast and tailored for specific users taking full advantage of the probability distribution provided by the NextGen methodology.