气候与环境研究
氣候與環境研究
기후여배경연구
CLIMATIC AND ENVIRONMENTAL RESEARCH
2013年
3期
407-413
,共7页
江志红%杭月荷%刘冬%吴息%熊海星
江誌紅%杭月荷%劉鼕%吳息%熊海星
강지홍%항월하%류동%오식%웅해성
覆冰%极值%重建
覆冰%極值%重建
복빙%겁치%중건
Wire icing%Extreme value%Reconstruction
基于我国南方有覆冰数据记录的气象站冰厚年极值及同期气象要素观测资料,统计分析多种气象因子对覆冰年极值形成条件频次分布的影响,归纳出了最易于出现覆冰年极值的温度、风速和湿度条件。在此基础上,通过对西南地区威宁、金佛山、峨眉山和三穗4站覆冰年极值与其相应气象变量的进一步分析,建立了覆冰极值序列的回归模型。根据现有气象站电线结冰资料及其对应时段的常规气象要素资料,对气象站电线结冰年极值序列进行重建试验,试验结果表明不同气候背景下覆冰极值序列的回归模型有显著差异。独立样本的交叉检验结果显示,威宁站年极值序列的回归模型效果较理想,重建序列能够较好地模拟覆冰的极值序列。
基于我國南方有覆冰數據記錄的氣象站冰厚年極值及同期氣象要素觀測資料,統計分析多種氣象因子對覆冰年極值形成條件頻次分佈的影響,歸納齣瞭最易于齣現覆冰年極值的溫度、風速和濕度條件。在此基礎上,通過對西南地區威寧、金彿山、峨眉山和三穗4站覆冰年極值與其相應氣象變量的進一步分析,建立瞭覆冰極值序列的迴歸模型。根據現有氣象站電線結冰資料及其對應時段的常規氣象要素資料,對氣象站電線結冰年極值序列進行重建試驗,試驗結果錶明不同氣候揹景下覆冰極值序列的迴歸模型有顯著差異。獨立樣本的交扠檢驗結果顯示,威寧站年極值序列的迴歸模型效果較理想,重建序列能夠較好地模擬覆冰的極值序列。
기우아국남방유복빙수거기록적기상참빙후년겁치급동기기상요소관측자료,통계분석다충기상인자대복빙년겁치형성조건빈차분포적영향,귀납출료최역우출현복빙년겁치적온도、풍속화습도조건。재차기출상,통과대서남지구위저、금불산、아미산화삼수4참복빙년겁치여기상응기상변량적진일보분석,건립료복빙겁치서렬적회귀모형。근거현유기상참전선결빙자료급기대응시단적상규기상요소자료,대기상참전선결빙년겁치서렬진행중건시험,시험결과표명불동기후배경하복빙겁치서렬적회귀모형유현저차이。독립양본적교차검험결과현시,위저참년겁치서렬적회귀모형효과교이상,중건서렬능구교호지모의복빙적겁치서렬。
The authors analyzed various factors that cause extreme ice covering (EIC) using statistical methods on extreme ice thickness records and contemporaneous observational meteorological data from weather stations in southern China, and summarized how certain meteorological conditions, such as temperature, wind speed, and humidity, affect the EIC value. Further study of the EIC values and meteorological parameters at stations of Weining, Jinfoshan, Emeishan, and Sansui, allowed us to establish a regression model for the EIC data series. Based on the extreme ice thickness records and contemporaneous observational meteorological data from weather stations, the wire EIC value sequences were reconstructed, and it showed that regression models of wire EIC value sequences under different climate background conditions have significant differences. According to cross-test results using independent samples, the reconstructed regression model sequence for the Weining station matched the observational data, and the model simulated the EIC sequence well.