Search

The spatiotemporal variability of droughts over a region can often be characterized by combining orthogonal climate modes with corresponding temporal coefficients. The temporal characteristics of climate modes have been extensively addressed, while their spatial development remains largely unexplored. Hence, this study investigated the spatiotemporal evolution of meteorological droughts from the perspective of climate modes. Nearly 50 years of monthly precipitation records (1960–2007) from 73 meteorological stations in the Poyang Lake basin, China, were used. The standardized precipitation index at a 1-month scale was applied to quantify meteorological droughts. The dominant modes of the droughts over the basin were identified using principal component analysis, K-means cluster analysis, and analysis of variance. Based on the trajectory migration identification method, the role of the climate modes in spatiotemporal evolution was analyzed. The results showed that four spatial modes of the droughts in the basin were identified. The spatial extents, centroids, and severity of the drought clusters based on the identified modes were significantly consistent with those based on the meteorological stations (R2 > 0.8, p < 0.05), which indicated that these modes could mostly match the large areas where serious dry/wet conditions occurred. Moreover, their performances in characterizing the spatiotemporal evolutional features (severity, migration distance, and pattern) of drought events were valid, which indicated that they might be considered as the elementary constituents of the historical meteorological drought events across the basin. The findings might offer some implications to understand drought development and causes through possible connections between the dominant modes and climate variability. [ABSTRACT FROM AUTHOR]

• A method to measure the spatial extent of drought at any location was proposed. • The parameters of the method were reasonably specified using sensitivity tests. • The method was tested over a humid region located in southern China. • The difference in spatial lengths of 8 directions showed the anisotropy of drought. • The spatial extents of multi-scale droughts decreased as the time scales increased. The spatial characteristics of droughts are important for understanding drought development and water resource allocation. Most previous studies mainly focus on the spatial display of drought characteristics and the identification of drought spatial extents. Little attention is given to measuring the spatial extent of drought at any location. To this end, a new framework was proposed for measuring the spatial areal extent of drought based on spatial autocorrelation and anisotropy principle. The Global Precipitation Climatology Centre monthly gridded precipitation data with a spatial resolution of 25 km × 25 km were used. The standardized precipitation index on a three-month scale (SPI-3) was applied to characterize meteorological droughts. The base period of 1961–2010 (50 years) was selected for calculating climatology lengths (Cls)/spatial extents (SPEs) of meteorological droughts over the Poyang Lake basin. The parameters of the proposed method, including the drought threshold (SPI ≤ −1), the number of search directions (8), search radius (500 km), distance step (25 km), bandwidth (56.95 km), and angle tolerance (22.5°), were examined using sensitivity tests. The results showed that notable differences in Cls of 8 search directions confirmed the spatial anisotropy of droughts. The statistics of SPEs of the SPI-3 series over the basin were reported. SPEs were mainly between 50 and 100 × 104 km2. The distinctive spatial pattern of the SPEs was revealed and characterized by a high-value belt from the northeast and southwest and two low-value areas. The SPI on other time scales (SPI-1/SPI-6/SPI-12) revealed similar characteristics as SPI-3. Moreover, SPEs are potentially associated with time scales, revealing a decreasing change (or increasing spatial fragmentation) as time scales increased. The proposed method based on anisotropy is reasonable and useful for mining the spatial characteristics of droughts and can be applied in other fields. [ABSTRACT FROM AUTHOR]