Future daily and annual thermal comfort of dairy cows and associated uncertainty

The thermal comfort of dairy cows was evaluated on the basis of the calculation of the temperature-humidity index (THI, calculated in 8 ways described in Dikmen and Hansen, 2009) and the attribution of a comfort class (no stress, mild stress, moderate stress, severe stress, emergency) according to the THI value and Collier et al. (2012) thresholds. To compute daily THI (provided in THI files), we used climatic data from two climate simulation exercises: Jouzel simulations (2014) and DRIAS simulations (2020), downloaded from the SICLIMA portal where only simulations from French laboratories are available. These climate simulation exercises provide daily climate variables from 1950 to 2100, on a spatial grid of 8 km side. In total, twelve climate simulations were considered, which correspond to the crossing of different global climate models (GCM), regional climate models (RCM) and representative concentration pathways (RCP). Daily average temperature and relative humidity were downloaded for four 30-year time horizons defined as: 'Reference': 1976-2005, 'H1': 2021-2050, 'H2': 2041-2070 and 'H3': 2071-2100. The indicator we used is the number of days per year that the animals are in each thermal comfort class (provided in D1 files). These number of days were then aggregated by simulation exercice, SAFRAN grid point, RCP and time horizon. Three types of aggregation (provided in D2 files) have been calculated corresponding to the median and the 5th and 95th percentiles of the distribution of these numbers of days according to the GCM and RCM of the simulation exercise and the 30 years of the considered time horizon. Maps (provided in D3 files) were generated for each of these aggregations and a shiny application was developed to allow the visualization of these maps. It is stored in the associated dataset "A shiny application to visualize the evolution of thermal comfort of dairy cows under climate change".