摘要
以咸阳“旱腰带”地区13个气象站2010—2019年气象监测资料为基础,通过分析该地区降水、气温及干旱的特征,结合1960年以来县气象站资料,模拟“旱腰带”地区近60年降水量,构建“旱腰带”地区降水量数据集;基于长短期记忆网络(LSTM),以历史降水量数据为输入对“旱腰带”地区的降水量进行预测;根据地区干旱分级标准,由预测得到的降水量结果计算得到不同地区的干旱级别,从而实现对干旱发展趋势的分析。结果表明,基于降水量预测数据得到的干旱分级与真实情况相比,精度达到85%以上,能够为“旱腰带”地区环境乃至旱区改善及生态修复提供理论基础和决策依据。
Based on the meteorological monitoring data of 13 meteorological stations in the“drought belt”area of Xianyang from 2010 to 2019,by analyzing the characteristics of precipitation,temperature and drought in this area,combined with data from county meteorological stations since 1960,the precipitation in the“drought belt”area in the past 60 years was simulated,and the precipitation data set in the“drought belt”area was constructed.Based on the long short-term memory(LSTM)network,the precipitation in the“drought belt”region was predicted with historical precipitation data as input.According to the regional drought classification standard,the drought level of different regions could be calculated from the predicted precipitation results,so as to realize the analysis of the drought development trend.The results showed that the drought classification based on the precipitation forecast data had an accuracy of 85%compared with the real situation,which could provide a theoretical basis and decision-making basis for environmental improvement and ecological restoration in the“drought belt”area.
作者
马筛艳
王薇
裴莉莉
吴玉龙
刘艳
MA Shai-yan;WANG Wei;PEI Li-li(Xianyang Institute of Agricultural Meteorology,Xianyang,Shaanxi 712034;School of Information Engineering,Chang’an University,Xi’an,Shaanxi 710064)
出处
《安徽农业科学》
CAS
2023年第13期192-197,共6页
Journal of Anhui Agricultural Sciences
基金
咸阳市重大科技专项“东庄水库区域生态环境变化与种植布局调整气象分析研究”(2020K01-33)。
关键词
干旱发展趋势
干旱等级
降水量预测
致旱原因
LSTM
Drought development trend
Drought level
Precipitation forecast
Causes of drought
LSTM
