Vignette

Climate BSC Group: Earth Sciences Software
A vignette which provides an example of code written in R that can be used to calibrate seasonal forecasts using several calibration methods. For each calibration method, different skill scores are calculated, compared to the ones obtained by using a non-calibrated forecast and finally displayed.
Software Author: 

Albert Martínez-Botí, Paloma Trascasa-Castro, An-Chi Ho, Raül Marcos-Matamoros, Carlos Delgado-Torres. Lluis Palma, Nuria Perez-Zanon, Alba Llabrés-Brustenga, Albert Vila Miró

License: 

GPL License (Version 3.0)

Primary tabs

The vignette shows an example conducted over a south-african domain, along a reference period going from 1995 to 2015 and for the variable "t2m" (air temperature 2m above the ground). The initialisation month of the hindcast — which is a forecast initialised in the past (known as well as retrospective forecast) — is November and the analysis is done for the seasonal aggregation of November, December and January (NDJ). This vignette also offers the possibility of choosing a different variable, region, and reference period than those specified in the example and can be run virtually via binder or locally on a personal computer. Depending on the selected parameters and the location where the vignette is executed, it may also download and store seasonal climate data from the Climate Data Store - Copernicus (CDS).