Modeling of spatial relationships of water vapor pressure using spatial statistics Techniques

Introduction Survey of spatial relationships of environmental data take into accounts as one of themost important goals of spatial statistics for analyzing the spatial patterns and understanding the spatial dependencies. In this context, Exploratory Spatial Data Analysis(ESDA) could well to provide methods for distinguishing betweenspatial randomandnon-random patterns. Using the ESDA for analyzing the spatial autocorrelation of climatic elements is necessary to distinguish the changes and spatial distribution of them. The present research has aimed at explaining the use ofESDA for explaining the spatial patterns ofwater vapor pressureas one of the most important climatic parameters. Water vapor pressure plays a crucial role in climate system as an important feedback variable associated with the earth’s energy balance and hydrologic cycle. This climatic parameter has an important rolein explaining the climate change or changes in climatic parameters, because of 1) It is the main sourceof rainfall in allweathersystems, 2) It suppliesthe latent heatin this process and controls the heat in thetroposphere, 3) It is the booster of the storm's speed and 4) plays a major role in the dynamics of atmospheric circulation. So, determination and interpretation of the likely reasons of Water vapor pressure changes and its variability are vitally important for human as well as other living-beings.
Materials & Methods The study area, with about 360,200 km2 area, is located in the south and southwest of Iran and approximately between 25° 00'N and 34° 25'N latitudes and between 45° 38'E and 59° 17'E longitudes. Southern and southwestern parts of the study area are located beside of two massive sources of moisture, Persian Gulf and Oman Sea. The main mountain chain in the study area is Zagros that extends from the northwest to the southern part of study area. The Zagros mountain range is responsible for the major portion of rain-producing air masses that enter the region from the western and northwestern sides, with relatively high amounts of rainfall. In this study, water vapor pressure data in pixels (dimension of 9×9 km) in the time interval 1981-2010 were collected by the Iranian Meteorological data website (http://www.weather.ir).For Interpolation the water vapor pressure Kriging, Inverse Distance Weighting (IDW) and Radial Basis Functions (RBF) were tested and so after the error validation, the optimum method (Ordinary kriging with Gaussian method) was chosen. As regards the aim of this study, analyzing the spatial variability of WV in regional and local scale, the most important geographic features such as elevation, longitude, latitude, slope and aspect were chosen. Topography maps of the study area collected by the Geological Survey of Iran (http://www.gsi.ir). By mosaicking, georeferencing and editing these maps in Arc GIS 10.2 software, the Digital Elevation Model (DEM) by 10 Km cell size was derived and based on it, the geographic features are prepared.Moran's I,Anselin local Moran's I and LISAasESDA’sapproaches were used to analyze the spatial autocorrelation of water vapor pressure patterns based on climate parameters.
Results & Discussion According to the cross validation, it was cleared thatthe optimum method for interpolation of water vapor pressureis Ordinary kriging with Gaussian method. The results of Moran’s Istatistic showed that the water vapor pressure hasspatial structure and distributed in cluster patternin the South and SouthWest of Iran. The monthly’s surveys showed thatthe spatial autocorrelation of water vapor pressure in warm months are higher than cold months and therefore have a greater tendency to cluster. The results alsoshowedthat the water vapor pressure tending to disperse and non-clusterinspace in the South and SouthWest of Iran. The bivariate Moran's Istatistic for relation of water vapor pressure and longitude showed the strong and positive spatial autocorrelation and also clustered pattern.
Conclusion Survey the monthly spatial autocorrelation of water vapor pressure showed that the water vapor pressure in warm month more dealing with high spatial autocorrelation than cold months and more inclined to clustering. The existence of such situationin the most regions of the study area in the warm seasonreflects the Consistency and homogeneity in this season in relation to the other seasons. Maybe the main reason of this circumstances are the lack of non-diversification of input pressure systems in this season, climate uniformity and sustainability and effects of local systems. Over time, the water vapor pressure in the south and southwest of Iran has tended to be more dispersed and less clustering in space.The reason for this incident is not fully revealed but it may be attributed to topographic effects, changes in system positioning, land use changes, etc.Investigating the relationship between spatial distribution of water vapor pressure and geographic parameters showed that the relationship betweenwater vapor pressureand latitude,elevation and slope suggested adispersed and heterogeneousspatial distribution between them. The results of the bivariat erelationship betweenwater vapor pressureand aspect suggested a discontinuity and randomness relation.