Development of a Physically Based PDSI and Its Application for Assessing the Vegetation Response to Drought in Northern China

Jing Xu,Li-Liang Ren,Xiao-Hong Ruan,Xiao-Fan Liu,Fei Yuan

DOI: https://doi.org/10.1029/2011jd016807

2012-01-01

Journal of Geophysical Research Atmospheres

Abstract:A physically based Palmer Drought Severity Index (PB‐PDSI) was developed by replacing the two‐stage bucket water balance model in the original PDSI with the distributed hydrological model. The PB‐PDSI was tested and compared with Liu's PDSI, which is applied commonly in China. The results indicate that the monthly water balance computation of the PB‐PDSI is a more realistic approximation when utilizing remote sensing data that characterize the land surface and incorporating the two‐source PET (Potential EvapoTranspiration) method, the hybrid runoff model and a flow routing module on the basis of daily hydrologic simulation. The impact of the vegetation cover change on drought was also reflected appropriately by the PB‐PDSI. The spatial distribution of the PDSI over the whole basin by the PB‐PDSI allows for the exploration of the relationship between vegetation vigor and moisture availability. Pearson correlation analysis indicated that the NDVI (Normalized Difference Vegetation Index) is highly and positively correlated to the PB‐PDSI during the first half of the plant growing season and that the two are weakly correlated or uncorrelated during plant senescence. This result implies that the vegetation response to water availability depends on the plant growth stage. June was shown to be the “moisture‐sensitive period” for the vegetation in the Laohahe River catchment. When the seasonality effects are incorporated, an accurate estimation of the relationship between the NDVI and PB‐PDSI can be derived using the regression model with dummy variables. This suggests that the PB‐PDSI has the potential to elucidate the interaction mechanism between drought and the changing environment.