东北地区玉米春季渍害指标构建及时空分布特征

为厘清东北地区玉米春季渍害的评价指标及变化特征，该研究利用东北地区164个气象站点1981-2020年逐日气象资料和历史灾情资料，采用相关分析和逐步回归方法，分析玉米春季播种-出苗期降水和温度对渍害等级的影响效应，建立当量降水量和当量温度，在此基础上构建渍害等级指标，探讨东北地区玉米春季渍害的时空分布特征。结果表明：1）玉米春季渍害指标验证基本一致率达82%，能较好地反映东北地区玉米播种-出苗期渍害的实际受灾情况。2）1981-2020年东北地区玉米播种-出苗期渍害的发生存在明显空间差异，高发区集中在辽宁省东部、吉林省东南部以及黑龙江省三江平原地区，渍害频率最高可达45%，轻度渍害发生范围明显扩大；玉米不同程度渍害发生频率存在差异，表现为轻度>中度>重度。3）玉米渍害发生频率总体呈上升趋势，但上升趋势不显著，其中2001年、2002年、2004年为显著突变年。研究结果能够为揭示气候变化背景下东北地区玉米春季渍害灾变机制和时空演变规律提供理论依据。

Indicator construction and spatial distribution characteristics of maize spring waterlogging in Northeast China

An evaluation index of spring maize waterlogging was constructed to analyze the spatial and temporal distribution characteristics in Northeast China. Daily meteorological, maize growth period, and historical disaster data were collected from 164 meteorological stations from 1981 to 2020. Pearson correlation analysis was used to screen the key disaster-causing factors of spring maize sowing-emergence seedling waterlogging. An analysis was made on the effects of early precipitation and temperature on the level of waterlogging. The stepwise regression was utilized to construct the equivalent precipitation and equivalent temperature, particularly with the degree of waterlogging as the dependent variable, while the factors with a significant impact on the waterlogging as the independent variables. A normality test of equivalent precipitation and the equivalent temperature was carried out to construct the waterlogging grade index. The lower confidence limit of 95% confidence interval was taken as the index critical value of mild, moderate and severe waterlogging grade of spring maize. The occurrence of waterlogging in the sowing-emergence stage of maize was inverted to analyze the availability of the index, compared with the historical disaster data of waterlogging. The trend analysis and mutation test were performed on the M-K test at the maize sowing-emergence stage waterlogging from 1981 to 2020. The spatial distribution and interdecadal variation characteristics were obtained for the different degrees of waterlogging. The results show: 1) The precipitation in the current process of waterlogging and the cumulative precipitation from 1 to 10 days before the process were significantly positively correlated with the grade of waterlogging disaster. The average temperature in the current process of waterlogging and the average temperature from 1 to 5 days before waterlogging were significantly negatively correlated with the grade of waterlogging disaster. Therefore, the waterlogging index was constructed for the maize sowing-emergence stage. The basically consistent rate of index verification was 82%, indicating the excellent accuracy of the index. Consequently, a better representative was achieved in the actual disaster situation of maize waterlogging in this period. 2) There were outstanding spatial differences in the occurrence of waterlogging at the sowing-emergence stage of maize. Specifically, the main frequency was concentrated in the eastern region, whereas, there was a low frequency of waterlogging in the East Four Leagues. Among them, the highest frequency of waterlogging was distributed in the eastern part of Liaoning Province, the southeastern part of Jilin Province, and the Sanjiang Plain of Heilongjiang Province, with the highest frequency of waterlogging up to 45%. The range of mild waterlogging showed an expanding trend. But, there was no change in the range of other degrees of waterlogging. Great differences were found in the frequency of different degrees of waterlogging in maize. The frequency from the high to the low was mild > moderate > severe. 3) The frequency of maize waterlogging damage showed an overall upward trend, but the upward trend was not significant (not through the 0.05 significance test). Among them, the years 2001, 2002, and 2004 were the significant mutation years (through the 0.01 significance test). The finding can provide a theoretical basis to reveal the disaster mechanism and temporal and spatial evolution of spring maize waterlogging in Northeast China under the background of climate change. A reliable scheme can be achieved in the real-time monitoring of waterlogging.