Spatiotemporal analysis of dew days in China's Loess Plateau

Global warming due to greenhouse effect has altered meteorological variables such as temperature, relative humidity, rainfall and sunshine hours. The resulting change of these variables could have strong effects that threaten population, agriculture, environment, economy and industry. It could even affect global food security and supply/demand of water resources in the world. The Loess Plateau in North China is a semiarid and sub-humid climate region and is well-known for severe soil erosion, fragile ecological environment and sensitivity to climate change. Climate change will have a major impact on the ecological environment and agricultural ecosystems. Given the above, temporal and spatial distribution of meteorological elements for the Loess Plateau region has been analyzed. However, there was little information on dew days on the plateau. Dew Dew day was a key parameter of hydrologic cycle and plant disease prevention. Analysis of the spatial distribution and long-term temporal trends of dew days and the relatedness with climatic variables may provide the basis for plant disease prediction and prevention in local areas. In this study, dew day data from 52 meteorological stations for the period 1961-2010 were calculated using a model. The spatial distribution of seasonal and monthly dew days was interpolated by Kriging and the temporal trends of the days examined using trend-free pre-whitening (TFPW) and Sen's slope estimator. Correlation analysis explained the dew-day formation. The results showed that at monthly scale, dew days started in March and ended in November, with a monthly mean of 7 dew days. The maximum dew days were in the south, southeast and northwest of the Loess Plateau in September, with a range of 8-12 days. Analysis of dew days indicated significant positive trends for 5.77%-25.00% of the stations, with a variation of 0.02-0.15 d·a^-1 during the periods from August through November and June. Dew days with significant negative trends were found too, with the decrease in July and April by 0.02-0.09 d·a^-1 and for 7.68%-17.31% of the stations. At seasonal scale, dew days occurred in spring, summer and autumn, with a seasonal mean of 15 dew days. The maximum dew days were in autumn, with 14-26 dew days in the south, southeast and northwest of the plateau. Dew days with significant positive trends were observed in summer and autumn, which varied respectively by 0.09-0.25 d·a^-1 and 0.09-0.15 d·a^-1 for 3.85% and 5.77% of the stations. Dew days with significant negative trends were evident in spring, which varied by -0.34 to -0.07 d·a^-1 for 5.77% of the stations. Relative humidity and temperature had clear and dominant effects on the spatiotemporal trend of dew days. The study provided a quantitative basis for understanding dew day distribution and trend in the Loess Plateau under global climate change. It also provided a vital reference for future plant disease forecast, prevention and risk assessment.