摘要
利用博罗县2006—2023年0~320 cm各层深度日平均地温数据,分析地温的月际、年际变化特征及其与汛期降水的统计关系。结果表明,地温年际变化整体呈偏暖波动上升趋势,深层地温的最热月和最冷月较浅层略有滞后,浅层地温的月际波动幅度较大,而深层则相对平稳。进一步相关性分析表明,1月160 cm、320 cm地温,10月160 cm地温,以及11月和12月的地表日最高温与次年汛期降水量存在显著统计相关性,提示地温变化在一定程度上可反映区域热力状态的前期积累信号。基于上述显著性因子构建多元线性回归模型,用于预测次年汛期降水量,2007—2023年逐年回代检验结果表明,该模型在一定程度上能够反映次年汛期降水变化趋势,为汛期降水预测方法的改进提供参考。
Based on daily average soil temperature data from depths of 0-320 cm in Boluo County from 2006 to 2023,this study investigates the inter-monthly and inter-annual variation characteristics of soil temperature as well as its statistical relationship with flood season precipitation.The results indicate that the inter-annual variation of soil temperature shows an overall warming trend with fluctuations.The hottest and coldest months of deep soil temperature lag slightly behind those of shallow layers,while the inter-monthly fluctuation amplitude of shallow soil temperature is larger,and that of deep layers is relatively stable.Further correlation analysis reveals a significant statistical correlation between soil temperatures at 160 cm and 320 cm in January,soil temperature at 160 cm in October,the daily maximum temperature of the ground in November and December,and the subsequent year's flood season precipitation.This suggests that changes in soil temperature can reflect the pre-accumulation signals of regional thermal states to some extent.A multiple linear regression model was constructed based on these significant factors to predict the subsequent year's flood season precipitation.The year-by-year back-test results from 2007 to 2023 show that the model can reflect the precipitation change trend in the next flood season to a certain extent,providing reference for the improvement of precipitation prediction methods in flood season.
出处
《科技创新与应用》
2026年第6期101-105,共5页
Technology Innovation and Application
基金
惠州市气象局科学技术研究项目(2023Q02)。
关键词
地温变化
降水预报
相关性分析
逐步回归
博罗县
ground temperature change
precipitation forecast
correlation analysis
stepwise regression
Boluo County
