Yosef Nabipour; Mahadi Vatakhah
Abstract
Rainfall spatial analysis methods are very helpful since there are not enough rainfall gauge stations and watersheds are scattered in large extent. There are many different methods for estimating average precipitation such as; arithmetic method and Thiessen polygon. However, the arrangement and location ...
Read More
Rainfall spatial analysis methods are very helpful since there are not enough rainfall gauge stations and watersheds are scattered in large extent. There are many different methods for estimating average precipitation such as; arithmetic method and Thiessen polygon. However, the arrangement and location of data and their correlations are not considered by classic methods. Thus, geostatistical techniques are applied instead. In the present article, 22 meteorological stations from within and around the basin with data collection period of 30 years were selected for the analysis. The geostatistic analysis methods including ordinary kriging, simple cokriging, ordinary cokriging, standardized ordinary kriging, moving average using inverse distance with powers of 1 to 5 were applied for spatial analysis of annual, monthly and 24 hourly maximum rainfall data in Hajighoshan watershed located in northeast of Iran. For this reason, rainfall data were fitted to different methods and compared using cross validation by removes rainfall values of each station, one at a time, and predicts the associated data value. The results of geostatistic analysis showed that ordinary kriging is the best method with MBE=34.26 for annual rainfall while moving average using inverse distance with power of 5 is the best method for monthly and 24 hourly maximum rainfall. According to the results obtained through analysis of variogram model, gaussian model are supposed as the best models for annual, monthly and 24 hourly maximum rainfall data.
Sharbanoo Abbasi Jondani; Ali Fathzadeh
Abstract
Snow is one of the main components of hydrological cycle in most of mountainous basins. Since collecting the snow data (e.g. snow water equivalent data) is very difficult and time consuming, some effort is necessary to develop methods to estimate spatially variation of snow depth distribution. In the ...
Read More
Snow is one of the main components of hydrological cycle in most of mountainous basins. Since collecting the snow data (e.g. snow water equivalent data) is very difficult and time consuming, some effort is necessary to develop methods to estimate spatially variation of snow depth distribution. In the present study, the at-site SWE data of 14 stations located in the west of Isfahan providence for the period 1989-2010 were spatialized applying four methods composing the Kriging, the Co-Kriging, the Radial Basis Functions (RBF) and the Inverse Distance Weighting (IDW). In order to reach this purpose, first, the normality of data was checked using the Kolmogorov – Smirnov test. The homogeneity, the stability and the trend of data were tested employing the semivariogram approach. Then the appropriate data of each year was entered into the ArcGIS 9.3 to conduct the methods. Finally, the best method for spatializing the SWE data was selected based on the RMSE values. The results showed that the RBF method provided the best results for most of the years. Furthermore, it was found that the amount of SWE reduced from the south and west to the north and east of the basin.
Hossein Eslami; Ali Salajagheh; Shahram Khalighi sigaroudi; hasan Ahmadi; Shamsollah Ayoubi
Abstract
Rainfall erosivity is the ability of rainfall to detach the soil particles. This study was conducted to evaluate spatial variability of rainfall erosivity indices in Khouzestan Province. The point data of indices (EI30, AIm, KE>1 and Onchev indices) in 74 stations were used to generate spatial erosivity ...
Read More
Rainfall erosivity is the ability of rainfall to detach the soil particles. This study was conducted to evaluate spatial variability of rainfall erosivity indices in Khouzestan Province. The point data of indices (EI30, AIm, KE>1 and Onchev indices) in 74 stations were used to generate spatial erosivity maps through deterministic and geostatistical interpolation methods (Radial Basis Functions, Inverse Distance Weighted, Kriging and Cokriging). Results indicate that cokriging have least error and most correlation with determining coefficient of 0.89, 0.89, 0.48 and 0.49 for EI30, AIm, KE>1 and Onchev indices. Based on the correlation relationships between the basins specific sediment yield (in basins dominating the sedimentation stations) and mean indices of EI30, AIm, KE>1 and Onchev, EI30 index with correlation coefficient of 0.98 (P<0.01) is selected as the appropriate rainfall erosivity index. Based on the prepared map on the basis of Cokriging method with secondary variable of maximum mean monthly rainfall, the east and northeastern regions presented the highest values of EI30 index, while the southern and western regions showed the lowest values of EI30 index. The annual rainfall erosivity (EI30) ranged from 404 to 3064 Mj.mm.ha-1.h-1.y-1.
