新疆参考作物腾发量多元回归模型及空间分布

Multi-variable regression models of reference crop evapotranspiration and the spatial distributions in Xinjiang

利用新疆境内42个气象站的多年气象资料,应用Penman-Monteith公式计算得到各站逐日和逐月尺度下的ET0值,应用多元统计回归分析得到相应的ET0与主要气象要素的回归方程,作为各地区计算ET0的经验公式.结合ArcGIS软件的空间分析功能,将各站回归方程的系数进行插值得出整个新疆地区ET0经验公式回归参数的空间分布图.结果表明,逐日和逐月尺度下,区域内的ET0空间分布差异显著,南部明显高于北部.回归所得的经验方程用来估算各地的ET0值是合理可靠的.回归所得的经验方程较Penman-Monteith公式简化得多,且具有一定精度,因此有一定的推广价值和应用前景.对于无气象资料地区,通过查图得出参数值,即可推算该区域特定时间尺度下的ET0值.

Meteorological dates of 42 weather stations in Xinjiang were used to estimate ET0 at daily and monthly time scales by using Penman-Monteith equation.Multiple-variable regressions of ET0 in terms of 4 major climatic elements were conducted to obtain empirical equations for calculating ET0.Combining with the spatial analysis function of code ArcGIS,the coefficients of the empirical regression formulas for whole Xinjiang were interpolated based on the regression formulas in 42 stations;consequently the spatial distribution of those coefficients was obtained.The results indicated that the spatial distributions of ET0 were obviously varied across Xinjiang both at daily and monthly scales.And the values of ET0 in southern part of Xinjiang were significantly larger than those in northern.It was reasonable and reliable to use the multiple-variable regression formulas to estimate ET0 in different parts of Xinjiang.Because the empirical formulas are simpler than Penman-Monteith equation and with relatively high accuracy,they have fairly promoting values and good application prospects.For the regions without meteorological data,by using the coefficient plots proposed in the paper,ET0 values in those regions can be estimated at a given time scale.