首页|宁夏历史降雨时空分布特征及基于MLP神经网络的趋势预测

宁夏历史降雨时空分布特征及基于MLP神经网络的趋势预测

Analysis of Spatial and Temporal Distribution Characteristics of Historical Rainfall in Ningxia and Trend Prediction Based on MLP Neural Network

扫码查看
为揭示宁夏历史降水及其变化趋势,通过滑动平均法、一次线性倾向法、降水集中法、超定量法、反距离权重插值法和MLP神经网络方法,对宁夏1955-2019年12个气象站点的降水数据进行处理和分析.结果表明,宁夏年降水量由南向北递减,年际降水量呈波动变化,有缓慢增长的趋势;极端降水事件多发生在南部黄土丘陵区;从预测结果看,宁夏2020-2025年进入降水枯水期的可能性较小,线性滑动平均法分析的降水量和MLP神经网络预测的降水量均达到多年平均水平.
To reveal the historical precipitation changes and trends in Ningxia region,the article used sliding average method,primary linear tendency method,precipitation concentration method,super-quantitative method,inverse distance weight interpolation method and MLP neural network to process and analyze the precipitation data from 12 meteorological stations in Ningxia region during 1955-2019.The results show that annual precipitation in Ningxia decreases from south to north,with fluctuating inter-annual precipitation and a slow increasing trend in terms of changes in average precipitation over the years;Most of the extreme precipitation events occurs in the Southern Loess Hilly Region;From the prediction results,it is less likely that Ningxia will enter the precipitation dry period between 2020 and 2025,and the precipitation after linear sliding and the precipitation predicted by MLP neural network both reach the multi-year average precipitation value.

precipitationspatiotemporal changeprecipitation predictionextreme precipitationconcentration index

赵逸雪、褚阳、张汉辰

展开 >

宁夏大学地理科学与规划学院,宁夏银川 750021

降水 时空变化 降水预测 极端降水 集中指数

宁夏回族自治区重点研发引才专项

2020BEB04027

2023

宁夏大学学报(自然科学版)
宁夏大学

宁夏大学学报(自然科学版)

CSTPCD
影响因子:0.377
ISSN:0253-2328
年,卷(期):2023.44(4)
  • 1
  • 18