Development of climate change scenarios for Latvia for the period until the year 2100

This report examines climatic changes projected for Latvia during the 21st century. Climate projections are based on a wide ensemble of state-of-the-art CMIP5 global climate models; that set of models was utilized in compiling the 5th Assessment Report of the Intergovernmental Panel on Climate Change. Projections have been elaborated separately for three greenhouse gas cenarios, the RCP2.6 scenario representing small, RCP4.5 medium and RCP8.5 large emissions. By the late 21st century, the following changes (expressed relative to the mean of the period 1971-2000) are projected: • In winter, mean temperatures are projected to increase by 1-4 °C under RCP2.6, 2-6 °C under RCP4.5 and 4-9 °C under RCP8.5. In summer, anticipated warming is weaker: 1-3 °C under RCP2.6, 1-4 °C under RCP4.5 and 2-7 °C under RCP8.5. In winter, warming appears to be somewhat larger in the eastern part of the country while in summer the geographical differences are small. • Diurnal temperature range would diminish in winter by 0-50 % and incident solar radiation by 0-30 %. In summer, changes in these quantities are most likely positive but fairly small. • Mean winter precipitation increases by 0-20 % under RCP2.6, 0-30 % under RCP4.5 and 10-50 % under RCP8.5. In summer, the sign of change is uncertain, but in southern Latvia it is somewhat more likely that precipitation decreases slightly rather than increases. • Ice days (with a maximum temperature below zero) become substantially less frequent while the count of summer days (maximum temperature above 25 °C) increases. • Thermal growing season would lengthen by up to two months and the degree day sum would nearly double (under RCP8.5). • According to the best estimate (multi-model mean), wind speeds would remain nearly unchanged throughout the year. Even so, scatter among the individual model projections is large, and in winter even changes larger than ± 20% are possible. When studying projections for a less distant future, the sign of change is the same as what is projected for the late 21st century but the magnitude is smaller. The above uncertainty intervals of projected changes reflect mainly the inter-model differences but the contribution of natural unforced variability has also been taken into account. In the course of the project, several data files have been delivered into Latvia to be used for additional analyses. These files include, for instance, the time series of 30-year running monthly mean changes and bias-corrected daily model output, both given for five climate variables and represented on a 10 x 10 km grid.