Uncertainty assessment of GCM models for estimating rainfall and runoff of Dez Ulya basin under climate change

Increase greenhouse gases in the Earth's atmosphere have led to imbalances in the phenomenon of climate over the past decades, defined as Climate Change. Studies show that climate change can have negative effects on water resources, agriculture, environment, health, industry and economy. Global warming and climate change is happening and changing weather and climate volatility is associated with greater risk of damage. Since increasing the likelihood of future climate change could have devastating consequences for human societies, it is essential to examine the drought situation in the future periods in this area. For climate change effects on various resources in the future, climatic variables affected by greenhouse gases should be determined. Different techniques are available to simulate the future climatic variables under climate change effects; the most reliable data is atmospheric general circulation models. GCM models are three-dimensional models of the physical relationships that govern the atmosphere, crysphere, biosphere and hydrosphere. One of the weaknesses of GCM models is large spatial and temporal scales of the climatic variables. Therefore variables regarding hydrological and water resources studies are not sufficiently accurate. It should be downscaled by various techniques. Since different methods are available for downscaling, the uncertainty associated with these methods must be investigated. Various uncertainties affect the final outcome runoff simulation in a basin under the impact of climate change. The credibility of the results by ignoring any of these uncertainties would be reduced.
In this study, the GCM models uncertainty, methods of downscaling climate models and the SRES emission scenarios over the period 2069-2040 on Dez Ulya basin runoff were examined. For this purpose, the simulated temperature and precipitation of 10 GCM models, including BCM2.0, CGCM3T63, CNRMCM3, CSIROMK3.0, GFDLCM2.0, GISS-ER, HADCM3, INMCM3.0, IPSLCM4, MIROC3.2MEDRES, with two downscaling methods (Change factor and statistical using LARS-WG software) and three emission scenarios (A1B and A2 and B1) and artificial neural network model were used to simulate rainfall-runoff model. LARS-WG (Long Ashton Research Station Weather Generator) is a stochastic weather generator which can be used for the simulation of weather data at a single site, under both current and future climate conditions. These data are in the form of daily time-series for suitahle climate variables, namely, precipitation (mm), maximum and minimum temperature (°C) and solar radiation (MJm-2day-1). Stochastic weather generators were originally developed for two main purposes: 1) To provide means of simulating synthetic weather time-series with statistical characteristics corresponding to the observed statistics at a site, but which were long enough to be used in an assessment of risk in hydrological or agricultural applications.2) To provide means of extending the simulation of weather time-series to unobserved locations, through the interpolation of the weather generator parameters obtained from running the models at neighboring sites. It is worth noting that a stochastic weather generator is not a predictive tool that can be used in weather forecasting, but is simply a means of generating time-series of synthetic weather statistically ‘identical’ to the observations. New interest in local stochastic weather simulation has arisen as a result of climate change studies. At present, output from global climate models (GCMs) is of insufficient spatial and temporal resolution and reliability to be used directly in impact models. A stochastic weather generator, however, can serve as a computationally inexpensive tool to produce multiple-year climate change scenarios at the daily time scale which incorporate changes in both mean climate and in climate variability. It utilizes semi-empirical distributions for the lengths of wet and dry day series, daily precipitation and daily solar radiation. The rainfall-runoff models for the base period (2000-1971) has been calibrated and verified, then by downscaling of ten GCM-AR4 climate models for the study area and take into account each of them separately for rainfall-runoff models, changes of runoff in the period 2069-2040 under the three scenarios (A1B and A2 and B1) were determined.
Results from downscaling models showed that the rainfall for some models increase and others decrease in the future, compared to the base periods. Changing factors in downscaling method showed more decrease than statistical method. Results showed that the percentage change in long-term monthly simulated runoff for the two downscaling methods is about 5.11 percent, while a decreasing trend in the future compared to the base runoff was seen. Runoff simulation scenarios relative to each other in different months had the same difference. The results showed uncertainty in climate models used in this study is more than of uncertainty according to downscaling methods and emission scenarios.