基于水文模型和遥感的干旱评估和重建

Drought evaluation and reconstruction based on a hydrologic model and remote sensing

为了描述干旱的空间及时间演进过程,并对比土壤水分干旱和植被干旱之间的异同,利用VIC(variable infiltration capacity)水文模型模拟的土壤水分数据及遥感的归一化植被指数(NDVI)产品,基于SAD(severity-area-duration)方法重建了Colorado流域2000—2008年的干旱事件,通过绘制SAD曲线分析了该流域的干旱特征,并比较了不同干旱指标的差异。结果表明:NDVI百分位数(NDVIP)和土壤水分百分位数(SMP)都可以用于重建Colorado流域的主要干旱事件,并描述对应干旱的特征;NDIVP监测的干旱事件持续时间小于SMP监测的结果。NDVIP具有反应干旱和湿润的能力,但其反映长期干旱的能力受植被生长周期的限制。

Droughts have spatial and temporal variations.Soil moisture data from the VIC（variable infiltration capacity）model and NDVI（normalized difference vegetation index）data from MODIS（Moderate Resolution Imaging Spectroradiometer）was analyzed using the SAD（severity-area-duration）method to investigate the soil moisture differences between natural droughts and agricultural droughts.The model was used to reconstruct and evaluate the 2000 to 2008droughts in the Colorado River Basin.The results indicate that the NDVIP（normalized difference vegetation index percentile） and SMP（soil moisture percentile）models can both be used for drought reconstruction with the durations of droughts detected by NDVIP tending to be shorter than those predicted by SMP.NDVIP has better potential for detecting drought and wetness,though the predictions are limited by the life cycle of the vegetation.