Summer Precipitation And Temperature Prediction In China Based On Soil Moisture And Year-To-Year Increment Approach

Based on the monthly precipitation (temperature) at 160 stations in China and the monthly surface soil moisture from ECMWF ERA-Interim reanalysis, the present study has selected year-to-year increment of soil moisture of different key regions over Eurasia as the predictors through correlation analysis, the statistical prediction models are then developed based on BP-CCA combining ECC method, to predict the year-to-year increment of summer precipitation (temperature) over Eastern China, and thus obtain the prediction of summer precipitation (temperature). Specifically, data during the epoch of 1980-2004 and 2005-2014 are used to perform the historical prediction test and the independent sample test, respectively. The main conclusions are as follows:(1) According to correlation analysis between the year-to-year increment of summer precipitation (temperature) over Eastern China and the year-to-year increment of April (May) surface soil moisture over Eurasia, we select year-to-year increment of soil moisture of 9 key regions as the predictors, to build single factor prediction models of the year-to-year increment of summer precipitation (temperature) with the BP-CCA method, and we calculate the predictive value of summer precipitation (temperature). It shows that the soil moisture of different regions have a certain instruction significance for the prediction of summer precipitation (temperature) and the year-to-year increment of summer precipitation (temperature) in China.(2) In the test of precipitation prediction, we develop ensemble prediction models with the ECC method based on different combinations of the 9 predictors, and the predictive effect are also evaluated, and then four optimal regional prediction models, which are prediction models of summer precipitation for North China, the middle and lower reaches of the Yellow River Basin, the Yangtze-Huaihe Region, South China, are selected. The soil moisture of the region S1 (the Eastern European Plain), S2 (the north of the Baikal Lake). S8 (Hetao Region), S9 (the south of the Yangtze River) are selected as the predictors for the prediction model of summer precipitation for North China. The soil moisture of the region S2 (the north of the Baikal Lake), S3 (the north of the Balkhash Lake), S4 (Northwest China), S8 (Hetao Region) are selected as the predictors for the prediction model of summer precipitation for the middle and lower reaches of the Yellow River Basin. The soil moisture of the region S3 (the north of the Balkhash Lake), S4 (Northwest China). S8 (Hetao Region), S9 (the south of the Yangtze River) are selected as the predictors for the prediction model of summer precipitation for the Yangtze-Huaihe Region. The soil moisture of the region S1 (the Eastern European Plain). S3 (the north of the Balkhash Lake), S8 (Hetao Region) are selected as the predictors for the prediction model of summer precipitation for the South China. In the independent sample test, the predicted precipitation trends obtained by the four models are consistent with observations in the corresponding regions. Prediction scores (PS) that are commonly used in the operational prediction of China all exceed 70 points, and anomaly correlation coefficient (ACC) that are generally used in the world are all positive. It shows that soil moisture contained useful signals for summer precipitation in China, and soil moisture can be considered to apply in the summer precipitation prediction operations.(3) In the test of temperature prediction, three optimal regional prediction models, which are prediction models of summer temperature for North China, the middle and lower reaches of the Yellow River Basin, the middle and lower reaches of the Yangtze River, are selected with the same way. The soil moisture of the region H1 (the lower reaches of the Lena River), H2 (the south of the Yellow River), H4 (the lower reaches of the Yenisei River), H5 (the West Siberian Plain), H6 (the northwest of the Indian Peninsula) are selected as the predictors for the prediction model of summer temperature for North China. The soil moisture of the region H1 (the lower reaches of the Lena River), H4 (the lower reaches of the Yenisei River), H6 (the northwest of the Indian Peninsula), H7 (the northeast of the Baikal Lake) the are selected as the predictors for the prediction model of summer temperature for the middle and lower reaches of the Yellow River Basin. The soil moisture of the region H2 (the south of the Yellow River), H4 (the lower reaches of the Yenisei River), H6 (the northwest of the Indian Peninsula), H7 (the northeast of the Baikal Lake), H8 (the west of the Baikal Lake) are selected as the predictors for the prediction model of summer temperature for the middle and lower reaches of the Yangtze River. In the independent sample test, the three prediction models all show good predictive skills. The rate of coherence of the temperature anomaly that the temperature prediction model over North China predicted is 8/10. The rate of coherence of the temperature anomaly that the temperature prediction model over the middle and lower reaches of the Yellow River Basin predicted is 9/10. The rate of coherence of the temperature anomaly that the temperature prediction model over the middle and lower reaches of the Yangtze River predicted is 7/10. By calculating the climatic scores of summer temperature over North China, the middle and lower reaches of the Yellow River Basin, the middle and lower reaches of the Yangtze River, we found that prediction scores (PS) all exceed 80 points, and anomaly correlation coefficient (ACC) all exceed 0.3. It shows that soil moisture contained useful signals for summer temperature in China, and soil moisture can be considered to apply in the summer temperature prediction operations.