Subjects: Geosciences >> Atmospheric Sciences submitted time 2023-08-25 Cooperative journals: 《干旱区研究》
Abstract： Based on the ECMWF fine grid numerical prediction product and the temperature observation data ofthe national assessment station, the weighted quasi- symmetric running training period method (WQSRTP) wasused to generate the maximum (low) objective product of the smart grid in Gansu Province. The results werecompared with the smart grid guidance forecast product (SCMOC) of China Meteorological Administration andthe urban grid forecast product (SPCC) of Gansu Province. The results show that the WQSRTP correction methodcan significantly improve the ability to predict the 24 h maximum (low) temperature of the ECMWF fine gridnumerical model, and the predictive accuracy of the 24 h maximum and minimum temperatures increased by32.16% and 15.48% , respectively. Compared with SCMOC, SPCC, and ECMWF, the modified WQSRTPproducts are positive correction techniques, and the modified ability of maximum temperature is better than thatof minimum temperature. According to the spatial error test, the WQSRTP correction method can effectivelyimprove the accuracy of maximum (low) temperature forecast in the Qilian Mountains and the SouthwestMountains, and significantly reduce the mean absolute error. Moreover, the effect of correction for predicting themaximum temperature is better than that of minimum temperature.
Hits
Hits
Downloads
Comment