| ABSTRACT | Precipitation, a crucial climate parameter, is in high demand for the research and operational activities in many different sectors as well as disciplines. Various gridded global and regional precipitation data sets are available, but they have discrepancies resulting primarily from the differences in the number of stations and interpolation techniques used to create them. The primary objective of this research is to gain insight into how consistently these datasets represent the temporal and spatial distribution of precipitation in Turkey, which has a quite complex topography. The gridded datasets include those obtained from station and satellite measurements (e.g., CRU, APHRODITE, E-OBS, TRMM), and those produced by reanalysis (e.g., ERA-INTERIM) and climate models (e.g., CORDEX simulations). Different periods and areas are used in the analysis depending on the availability of the data. The preliminary results indicate that there are not only spatial but also temporal differences between observations both at country and basin scale. The discrepancy amongst observations is larger in the northern basins of the country that are more mountainous. In addition, the discrepancy is usually large in the winter and spring months. The models have a tendency to produce higher precipitation values than the observations in winter and spring. The major spatial difference between the model- and observation-based datasets takes place in the mountainous areas lacking an adequately dense station network. The most outstanding example is the northeastern Black Sea region where the observations show a decreasing precipitation with elevation while the models indicate an increasing precipitation, especially above a certain altitude. (This study, which is in progress, has been supported by TUBITAK (The Scientific and Technological Research Council of Turkey) under project number 114Y114.) |