Incoming solar radiation is the primary driver for physical and biological process in the earth. Human activities, such as
agriculture, forestry, land management, etc, ultimately depend on solar radiation. At a global scale, the geometry of earth's rotation and revolution about the sun cause the gradients of solar radiation. But topography is the major factor modifying the distribution of solar radiation at a landscape scale. Spatial solar radiation models provide a cost-efficient means for understanding the spatial variation of solar radiation over landscape scales. Geographic Information system (GIS) has become established tools for analyzing such models. Among such models, the Solar Analyst draws from the
strengths of both point-specific and area-based models. It can calculate solar radiation integrated for any time period. In
this paper, this model was used to estimate the spatial distribution of incoming potential solar radiation in Xinjiang, China. The 1km resolution digital elevation model (DEM) derived from 1:250000-scale topographic maps and other topographic factors (altitude, slope, aspect, etc derived from DEM) of Xinjiang were used as the basis for generating digital maps of the important parameter required to run Solar Analyst model. With the assistant of topographic factors
and Solar Analyst model, the spatial distribution of monthly incoming potential solar radiation with 1km resolution was estimated. Actual solar radiation data were obtained from 13 meteorological stations for the result validation. Validation determined that the mean relative error (MRE) of incoming potential solar radiation ranges from 3.8% in Jul to 12.2% in Dec and the mean value of monthly MRE is 7.1%. The MRE is larger in winter than in other season. In conclusion, the simulated results of model are basically up to the level of application requirement. The Solar Analyst model may serves
as a good tool for estimating spatial patters of monthly incoming potential solar radiation in Xinjiang. Application of Solar Analyst in Xinjiang and analysis of the spatial distribution characteristics of monthly incoming potential solar radiation have great significance for the research fields of agriculture, forestry and ecology in Xinjiang, China.
Potential evapotranspiration (PET) was estimated by applying Penman-Monteith Method recommended by FAO with climatic data from 96 stations during 1961 to 2000 in Xinjiang. The spatial and temporal variations of the potential evapotranspiration in Xinjiang are analyzed. The whole potential evapotranspiration has decreased in all seasons. The average annual evapotranspiration rate decreases by 29.87 mm/decade. Superimposed on this general decline are fluctuations with above average rates in the 1970s and 1980s. Decreasing PET rates are more pronounced in spring and summer as compared to autumn and winter. Through correlation analysis, the major climate factors that affect the temporal change of the potential
evapotranspiration are analyzed. Changes in relative humidity and to a lesser degree wind speed and sunshine duration were found to be the most important meteorological variables affecting PET trends in Xinjiang while changes in temperature played an insignificant role. Negative evapotranspiration trends are thought to a general decrease under global warming scenarios.
Precipitation is a function of many topographical features as well as geographical locations. The correlations between precipitation and topographical and geographical features can be used to improve estimation of precipitation distribution. In this paper, we built seasonal precipitation model based on GIS techniques in Zhejiang Province in southeastern China. Terrain variables derived from the 1 km resolution DEM are used as predictors of the seasonal precipitation, using a regression-based approach. Variables used for model development include: longitude, latitude, elevation, and distance from the nearest coast, direction to the nearest coast, slope, aspect, and the ratio of land to sea within given radii. Seasonal precipitation data, for the observation period 1971 to 2000, were assembled from 59 meteorological stations. Precipitation data from 52 meteorological stations were used to initialize the regression model. The data from the other 7 stations were retained for model validation. Seasonal precipitation surfaces were constructed using the regression equations, and refined by kriging the residuals from the regression model and subtracting the result from the predicted surface. Latitude, elevation and distance from the sea are found to be the most effective predictors of local seasonal precipitation. Validation determined that regression plus kriging predicts mean seasonal precipitation with a coefficient of determination (R2), between the estimated and observed values, of 0.546 (winter) and 0.895 (spring). A simple regression model without kriging yields less accurate results in all seasons.