Search

Loess Plateau has great uncertainty on drought occurrence due to climate change. This paper analyzes the evolution of precipitation, potential evapotranspiration and standardized precipitation evapotranspiration index (SPEI) based on the Coupled Model Inter-comparison Project Phase 5 (CMIP5) data and regional downscaling model (RegCM4.0). Results indicate that, under RCP2.6 Scenario, the precipitation will increase significantly (5% confidence level) at the rate of 16.40 mm/10 a. However, the potential evapotranspiration is showing non-significant decreasing trend at the rate of 2.16 mm/10 a. Moreover, the SPEI will decrease in the south and northernmost area and increase in the central northern area of Loess Plateau. Under RCP8.5 Scenario, the precipitation will increase significantly (5% confidence level) at the rate of 19.12 mm/10 a. The potential evapotranspiration will non-significantly decrease at the rate of 2.16 mm/10 a and the SPEI is showing increasing trend almost in the whole Loess Plateau. Generally, Loess Plateau is becoming wetter in the central part under RCP2.6 Scenario and the wet area will be enlarged to almost the whole plateau under RCP8.5 Scenario. Based on the results, the water resources will increase under global warming, which may alleviate the water scarcity issue in the Loess Plateau. [ABSTRACT FROM AUTHOR]

Summary: This paper develops tools to assess regional drought events based on limited periods of record and applies them to regional climate model output to examine potential future changes in drought due to increased greenhouse gases. The drought metric was defined as a monthly time series of standardized 12-month precipitation accumulations. Regional drought events were characterised by determining the severity, area, duration and frequency of dry periods. In order to improve statistical inference of drought events, a drought generator was adopted to synthesis a large number of events trained on the available data. In nearly all cases, the drought generator produces synthetic events with drought characteristics statistically indistinguishable from the training data. Severity–Area–Frequency (SAF) curves and Area–Probability (Area–P) curves could then be constructed from the synthetic data to determine the return period of a drought event of a given severity covering a given area. These techniques were applied to observations and to output from the Hadley Centre regional climate model (HadRM3). HadRM3, when driven by climate reanalysis data, has good correlation with the observed drought metric. HadRM3 also captures the timing of observed dry periods but not the relative seriousness of any particular period. Natural variability is large with regional drought characteristics highly dependent on the time period covered by the training data. Although the synthetic drought characteristics of HadRM3 generally overlap with the observed natural variability; the model slightly overestimates the area of observed drought; underestimate its frequency; and underestimate the severity at any given area. HadRM3 projections of changes in drought characteristics over the second half of the 21st century under increased atmospheric greenhouse gases hint at an increase in the severity of drought. However, any changes are not readily distinguishable from natural variability or projection uncertainty. Therefore it is not yet possible to robustly predict changes in UK meteorological droughts arising from increased greenhouse gasses. [Copyright &y& Elsevier]