Estimation and Analysis of Near Surface Vapor Pressure in China Based on MODIS Data
基于MODIS数据的中国地面水汽压模拟与分析

Combined daily infrared data with near-infrared data (MOD05) from the Moderate Resolution Imaging Spectroradiometer (MODIS) in year 2006, monthly precipitable water vapor was calculated using a weighted arithmetic. The monthly average precipitable water vapor ranged between 0.10 cm and 3.50 cm. With routinely observed meteorological data at 619 stations across Mainland China, an empirical model was build to estimate monthly near surface vapor pressure with the spatial resolution of 1 km × 1 km. Due to an extensive territory and different types of climates of China, an optimized method was employed to estimate the parameters of the empirical model. The main idea of the optimized method was to identify similar geographic coordinates and altitudes. The optimized parameters exhibit significant heterogeneity. Results are given as follows. 1) The empirical model appears to be capable of generating monthly near surface vapor pressure. Comparison between the simulations and observations of near surface vapor pressure at 33 stations across China indicates a correlation coefficient up to 0.97 and more than 90% samples with relative errors smaller than 20%. In particular, validation performed at 100 weather stations in Henan Province shows a correlation coefficient of 0.96 and the mean relative errors smaller than 20%, in which the mean relative errors in eight months were within 10%; 2) Across the entire country, the monthly average vapor pressure in 2006 ranged from 3.47 hPa to 17.13 hPa, with an annual mean of 8.87 hPa. The spatial variation varied between 3.20 and 8.66, with an annual mean of 5.36. The value and spatial variation of the monthly vapor pressure were higher in summer (Jun, Jul, and Aug) but lower in winter (Dec, Jan, and Feb). The maximum value occurred in Aug whereas the minimum value in Jan. A particular investigation into Yunnan Province, Southwest China, was performed. The spatial distribution of vapor pressure corresponded to DEM and river networks; 3) Distribution of the monthly vapor pressure varied markedly with the topography. Changes in vapor pressure with the altitude, slope, and aspect were analyzed to look at distributions of vapor pressure at local and regional scales. The monthly vapor pressure varied regularly with the topography in low altitudes. There were generally two peaks and two troughs in these vapor pressure values in different months, the same as in different altitudes. However, there existed a more complex characterization of the monthly vapor pressure at high altitudes such as over the Tibet Plateau. The proposed model for generating near surface vapor pressure based on MODIS data seems to be a promising tool in study of the energy balance and atmospheric water cycle under rugged terrains.