Air quality ensemble forecasting method and device

Abstract

The invention discloses an air quality ensemble forecasting method and device. The method comprises the following steps: acquiring multi-mode air quality forecasting data of a specified region; acquiring site air quality monitoring data of a specified area; for a plurality of integration algorithms, according to the multi-mode air quality forecast data in the training period and the corresponding station air quality monitoring data, respectively calculating parameter values of the integration algorithms, and obtaining the integration algorithms after training is completed; performing forecasting test on the same pollutant in the same area by using each trained integrated algorithm, comparing a forecasting test result with corresponding site monitoring data, and selecting an optimal integrated algorithm according to a comparison result of each integrated algorithm; and forecasting the air quality by using the selected optimal integration algorithm. According to the method, the computing resource demand and complexity of ensemble forecasting are reduced, and business application is facilitated; and an optimal set strategy is constructed through dynamic rolling test forecasting, and self-adaption and dynamic optimization of parameters of an integrated algorithm are achieved.

Description

technical field [0001] The present invention relates to air quality forecasting and early warning technology, and more specifically, the present invention relates to an air quality collective forecasting method and device. Background technique [0002] Air pollution has become a major environmental problem in cities today. Air quality forecasting is an important prerequisite for precise air pollution prevention and control and effective response to heavily polluted weather. In addition, reasonable and accurate air quality forecasts can effectively guide production activities and public travel, and reduce the adverse impact of pollution on production and life. [0003] The accuracy of the air quality model forecast is affected by many factors, and there are large uncertainties, including meteorological conditions, pollution emissions, initial conditions, boundary conditions, and the physical and chemical parameterization scheme mechanism of the model itself. How to continuo...

AI Technical Summary

Problems solved by technology

[0006] The present invention innovatively provides an air quality ensemble forecasting method and device, which solves the problem that the prior art cannot simultaneously obtain the optimal ensemble effect in different regions, pollutants, time periods and / or timeliness

Images