基于遥感技术的1988—2018年渤海海冰冰情时空变化特征

基于AVHRR、MODIS、GOCI遥感数据解译，构建了1988—2018年非连续日尺度冰情序列数据集，分析了渤海海冰冰情年际和年内时空变化特征。渤海冰情等级、年均海冰面积和年均海冰厚度均有明显的年际波动。海冰生消过程的日均面积和厚度变化呈现单峰型，且海冰面积和冰厚峰值均出现在1月25日。渤海海冰主要分布在辽东湾，其冬季结冰范围占渤海曾经结冰范围的70.58%,渤海中间区基本不结冰。海冰冰情指标与气象因子相关性分析结果表明：冻冰期海冰面积和冰厚相关系数最高的气象因子均是累积冻冰度日(CFDD);融冰期海冰面积相关系数最高的气象因子是3d-1.8℃积温，而冰厚则是累积融冰度日(CTDD)。据此选取CTDD、CFDD、3d-1.8℃积温及日最低气温四个指标建立多元非线性回归模型，并对冻冰期和融冰期分别拟合。进而对比分析2014—2018年海冰遥感提取面积与回归模型计算得到的海冰面积，其整体变化趋势一致，有一定的预测性。

Temporal and Spatial Variation Characteristics of Sea Ice Conditions in the Bohai Sea from 1988 to 2018 Based on Remote Sensing Technology

Based on the interpretation of AVHRR,MODIS,and GOCI remote sensing data,a discontinuous daily-scale ice sequence data set from 1988 to 2018 was constructed,and the characteristics of inter-annual and intra-annual temporal and spatial changes in the Bohai sea ice condition are analyzed.Bohai sea ice index,average annual sea ice area and annual average sea ice thickness all have obvious inter-annual fluctuations.The daily average area and thickness of the sea ice generation and disappearance process showed a single peak type,and the sea ice area and sea ice thickness’s peak all occurred on January 25.Bohai sea ice is mainly distributed in Liaodong Bay,its ice range accounted for 70.58 percent of the Bohai Sea once frozen range,Bohai’s middle area is basically non-frozen.The results of correlation analysis between sea ice indicators and meteorological factors show that the meteorological factor with the highest correlation coefficients of sea ice and ice thickness during freezing ice age is cumulative freezing degree days(CFDD),and the meteorological factor with the highest correlation coefficients of sea ice area during melting ice age is three-day-1.8℃accumulation temperature,while ice thickness is cumulative thawing degree days(CTDD).Based on this,four indicators of CTDD,CFDD,three-day-1.8℃accumulated temperature and daily minimum temperature were selected to establish a multiple nonlinear regression model,and the freezing and melting periods were respectively fitted.Furthermore,it compares and analyzes the sea ice remote sensing area extracted from 2014—2018 and the sea ice area calculated by the regression model.The overall change trend is consistent,which is predictive.