Based on the data of sandstorm at 11 stations in Ulanqab City from 1990 to 2021,the spatial and temporal variation characteristics of sand-storm weather were analyzed firstly,and then the conceptual models of cold fro...Based on the data of sandstorm at 11 stations in Ulanqab City from 1990 to 2021,the spatial and temporal variation characteristics of sand-storm weather were analyzed firstly,and then the conceptual models of cold front and Mongolian cyclone sandstorm were obtained by analyzing sandstorm cases.Finally,the forecast points of the two types of sandstorm weather were given to provide some scientific basis and reference for the prediction of local sandstorm weather in the future.展开更多
[Objective] The reason for the unsuccessful forecast of a heavy rainfall event in Yingkou was analyzed. [Method] Based on the precipitation data observed by automatic weather stations and MICAPS data, a heavy rainfall...[Objective] The reason for the unsuccessful forecast of a heavy rainfall event in Yingkou was analyzed. [Method] Based on the precipitation data observed by automatic weather stations and MICAPS data, a heavy rainfall Event was studied in Yingkou from 19 July to 21 July in 2010. Then the analysis of an unsuccessful forecasting for the heavy rainfall on 21 July was illustrated by CINRAD-SA data, satellite data and numerical forecast products. [Result] The main reason for the unsuccessful forecast was that the duration of the rainfall was long and inconsecutive. The distribution was uneven. Strong precipitation on 21st was different from the one in previous two durations. It was regional short term strong precipitation. And the forecast difficulty was large; the numerical forecast was unstable and erroneous;strong precipitation occurred in the night on 20th, which was shortly before the strong precipitation in the evening of 21st. This would easily confuse the reporter. Besides, the short term stillness of radar and cloud during this time would form certain disturbance. The focus of rainstorm forecast should based on the numerical forecast instead of element forecast;insisting on situation analysis and taking element judgment as auxiliary;as for strong precipitation forecast, there was large error in numerical forecast and can not be relied. Reporter should report the correct one based on experience. [Conclusion] The study provided reference for the forecast of rainstorm.展开更多
Time series data is a kind of data accumulated over time,which can describe the change of phenomenon.This kind of data reflects the degree of change of a certain thing or phenomenon.The existing technologies such as L...Time series data is a kind of data accumulated over time,which can describe the change of phenomenon.This kind of data reflects the degree of change of a certain thing or phenomenon.The existing technologies such as LSTM and ARIMA are better than convolutional neural network in time series prediction,but they are not enough to mine the periodicity of data.In this article,we perform periodic analysis on two types of time series data,select time metrics with high periodic characteristics,and propose a multi-scale prediction model based on the attention mechanism for the periodic trend of the data.A loss calculation method for traffic time series characteristics is proposed as well.Multiple experiments have been conducted on actual data sets.The experiments show that the method proposed in this paper has better performance than commonly used traffic prediction methods(ARIMA,LSTM,etc.)and 3%-5%increase on MAPE.展开更多
文摘Based on the data of sandstorm at 11 stations in Ulanqab City from 1990 to 2021,the spatial and temporal variation characteristics of sand-storm weather were analyzed firstly,and then the conceptual models of cold front and Mongolian cyclone sandstorm were obtained by analyzing sandstorm cases.Finally,the forecast points of the two types of sandstorm weather were given to provide some scientific basis and reference for the prediction of local sandstorm weather in the future.
文摘[Objective] The reason for the unsuccessful forecast of a heavy rainfall event in Yingkou was analyzed. [Method] Based on the precipitation data observed by automatic weather stations and MICAPS data, a heavy rainfall Event was studied in Yingkou from 19 July to 21 July in 2010. Then the analysis of an unsuccessful forecasting for the heavy rainfall on 21 July was illustrated by CINRAD-SA data, satellite data and numerical forecast products. [Result] The main reason for the unsuccessful forecast was that the duration of the rainfall was long and inconsecutive. The distribution was uneven. Strong precipitation on 21st was different from the one in previous two durations. It was regional short term strong precipitation. And the forecast difficulty was large; the numerical forecast was unstable and erroneous;strong precipitation occurred in the night on 20th, which was shortly before the strong precipitation in the evening of 21st. This would easily confuse the reporter. Besides, the short term stillness of radar and cloud during this time would form certain disturbance. The focus of rainstorm forecast should based on the numerical forecast instead of element forecast;insisting on situation analysis and taking element judgment as auxiliary;as for strong precipitation forecast, there was large error in numerical forecast and can not be relied. Reporter should report the correct one based on experience. [Conclusion] The study provided reference for the forecast of rainstorm.
基金This work is supported by the National Natural Science Foundation of China(Grant No.61971057)MoE-CMCC Artifical Intelligence Project(No.MCM20190701).
文摘Time series data is a kind of data accumulated over time,which can describe the change of phenomenon.This kind of data reflects the degree of change of a certain thing or phenomenon.The existing technologies such as LSTM and ARIMA are better than convolutional neural network in time series prediction,but they are not enough to mine the periodicity of data.In this article,we perform periodic analysis on two types of time series data,select time metrics with high periodic characteristics,and propose a multi-scale prediction model based on the attention mechanism for the periodic trend of the data.A loss calculation method for traffic time series characteristics is proposed as well.Multiple experiments have been conducted on actual data sets.The experiments show that the method proposed in this paper has better performance than commonly used traffic prediction methods(ARIMA,LSTM,etc.)and 3%-5%increase on MAPE.