【作者】 杨凤海；

【导师】 沈能展；

【作者基本信息】 东北农业大学 ， 作物栽培学与耕作学， 2010， 博士

【摘要】 近些年,气候变化不断突显,全球变暖、大气环流异常、极端气候频繁,世界各国对此十分关注与重视。黑龙江省是东北地区和全国气候变暖最大地区,极端天气不断增多,灾害繁发。因此,对气候变化分析、评价与监测工作手段和效率提出了更高要求,地理信息系统(GIS)等空间信息技术在气候资源时空变异研究中的应用越来越广泛和深入。GIS是“3S技术”(遥感RS、全球定位系统GPS和地理信息系统GIS)的核心,以GIS为支撑的空间信息技术在气候要素插值和气候变化分析中应用研究不断深入。为了探索黑龙江省气候变化的区域特征和规律,本文以GIS技术和地统计分析为支撑,将旬气候要素数据作为基础,通过相关模型对来自全省气象站点的气候资源数据进行插补,将插值得到的区域气候要素空间表面数据存贮在数据库中,利用站点数据和插值结果分析了全省气候资源时空变异规律。研究成果对于区域气候要素插补方法选择、气候资源数据库建设、区域气候变化分析与监测、提高区域气候变化和气候要素空间变异分析手段、完善农业信息化内容和手段、开发利用区域资源和农业发展决策等方面具有重要意义。具体研究成果如下：1.气候要素数据的空间插补与计算。本文以ArcGIS为支撑,80个气象站点观测的1997-2006年10年的气候要素数据为插值样本,考虑区域地面高程(DEM)、坡向、森林覆盖率、土地利用覆被等因素的影响,采用协同克里格(Cokriging)、径向基(RBF, Radial Basis Functions)等方法,对全省旬气候要素如气温、降水量、日照时数、蒸发量等进行空间插值,得到它们全省范围内1km×1km所有单元较高精度的空间表面数据,其中36个旬降水量插值结果均误差、平均标准差和均方根标准差的平均值分别为-0.007mm、-0.015和1.011,表明所用插值方法对于气候要素的插值是一种适当、可信的方法。进而通过地图代数方法(栅格叠加)得到相应气候要素月、年时间序列空间分布数据。2.气候要素空间数据库创建。将插补得到的气候要素及相关数据进行整理,以ArcGIS Geodatabase模型为支撑,在ArcCatolog环境下,建立了集旬、月和年时间序列的气温、降水量、积温、温润指数、地温、蒸发量等气候要素以及土地利用区域、行政区划、气象站点等栅格数据集(Raster Datasets)、矢量要素集(Feature Datasets)于一体的气候要素空间数据库。3.气候要素时空变异分析。通过气候要素的时空变异分析,得出全省主要气候要素的时空变化规律如下：1)近10a来,旬平均气温有波动。从空间上看,东南部和东部地区变化较小,其它地区变化较大；从时间上看,11-13、12-14、19-21旬的平均气温有平稳下降趋势,15-17、26-28和27-29旬的平均气温有平稳升高趋势：7月份平均气温有稍许下降趋势,9和11月的平均气温稍有上升趋势,5-9月平均气温升高约1.0℃,平原地区升温幅度大于山地丘陵区。气象站年平均气温移动变化以2.9℃为均值在2.5-3.3℃之间,略有升高和波动迹象,但无明显上升趋势。与多年平均气温相比,也无明显升高。春夏之交一些旬期平均气温变化率降低趋稳,夏秋之交一些旬期平均气温变化率升高,表明实际物候有向后延迟的迹象。2)旬降水量及其波动在时间上呈中间大、两头小的分布特征；月降水量也呈中间大、两头小,但其波动却是中间小、两头大；年降水量基本是以小兴安岭和张广才岭地区为最大,向周边逐步减少,且以向大庆、齐齐哈尔方向递减最快,其它方向递减较慢。原来多年降水量较大的峰值区域降水量减少,地区间降水量较差减小。全省插值和计算的年降水量为524.6mm,与历史数据相比减少幅度约为60mm。小兴安岭和张广才岭一带降水量相对较大,松嫩平原西部和大兴安岭西麓相对较小,地区间降水量分布不均。3)13-27旬湿润指数在平原地区波动变异较大,山地丘陵地区变异较小。齐齐哈尔变异最大,牡丹江变异最小。7月份波动变化最小,最大在春季的5月份。4)13-27旬的积温在山地丘陵地区和松嫩平原地区变异较大,三江平原地区变异较小。松嫩平原和三江平原这两大平原之间的积温变异存在较明显的差别。大兴安岭、伊春、齐齐哈尔等地、市旬积温波动变异较大,七台河、佳木斯等市波动变异较小。山地丘陵地区的地、市年积温相对较低,而位于平原地区和东南部地区相对较高。七台河、鸡西、佳木斯、双鸭山等市≥0℃年积温波动变异较小,大兴安岭、伊春、齐齐哈尔等地、市波动变异较大；七台河、佳木斯、鸡西等市≥10℃年积温波动变异较小,大兴安岭、伊春、齐齐哈尔、绥化等地、市波动变异较大。创新点：1)首次利用协同克里格(Cokriging)方法,建立黑龙江省气候要素与高程(DEM)、坡向、森林覆盖率等因素间的统计关系,对降水量、蒸发量、日照时数等气候要素进行空间插值。基于ArcGIS对气温、降水、积温、湿润指数等气候要素进行分析,得到它们的区域空间变异规律；2)在计算旬湿润指数过程中,深入研究了多种湿润指数计算的模型,参考了适合本论文研究的模型——德·马东(de Martonne)月湿润指数模型IdM=P/(T+10) (P为月降水量,T为月平均气温),吸收该模型用降水量与气温间关系反映湿润程度的思想,并通过相关分析寻找模型I=P/(T+a) (P为旬降水量,T为旬平均气温,a为常数)计算结果同P／E(P为旬降水量,E为旬蒸发量)间的关系,从而建立适合计算黑龙江省旬湿润指数的公式IYFH=P/(T+5) (P为旬降水量,T为旬平均气温)。用新创建的模型计算旬湿润指数,收到很好的效果；3)气候要素空间插值主要按旬进行,插值出旬平均气温、降水量、日照时数、蒸发量等与农业生产关系密切的气候数据,进而计算月和年的气候数据,并基于ArcGIS的Geodatabase模型建立相应的空间数据库,为作物自然生产潜力研究奠定基础,同时,也弥补了以往研究成果中旬气候要素空间分布数据的不足。更多还原

【Abstract】 In recent years, climate change, global warming, atmospheric circulation exceptions and extreme climate often appeared, attracting the world’s great care and attention. Heilongjiang Province is the region where air temperature were warming highest in northeast china also as the country, which extreme weather is growing and hazards increased. Therefore, there are higher demand to technologies and work efficiency on climate change analysis, evaluation and monitoring, while the spatial information technologies such as Geographical Information System (GIS) have been more and more applied in research on spatiotemporal variation of climate changes widely and deeply.GIS is the core of "3S technology" (Remote Sensing, RS; Global Positioning System, GPS and Geographic Information Systems, GIS). Spatial information technology have been applied in research on the interpolation of climate elements and its changes supported by GIS. In order to explore the regional characteristics of climate changes in Heilongjiang province, climate elements were interpolated in this paper with the support of GIS techniques and statistical analysis, based on 10-days data from wheather stations, through the relevant models, to get the spatial surface data stored in the database. Spatial and temporal variation laws of climatic resources all over the province was found by exporing and analysing data from weather stations and its interpolated values. The results of this research are significent to methods selection of climate element interpolation in a region, geodatabase established to store climage resources, analysis and monitoring of regional climate change, improving the means to analysis of regional climate change and climate element variations, promoting the content and means of agricultural informationization, development and utilization of regional resources and decision-making for agricultural development. The detail results of this research are as follows:1. The results in spatial interpolation of climate elements. In this paper, supported by ArcGIS, using climate elements data obseved in 80 weather stations in 1997-2006 as sample data to be interpolated, considering the influecne of the regional ground elevation (DEM), slope aspect, forest cover, land-use cover and other factors on climate elements, adopting Cokriging, Radial Basis Function methods, the province’s climate elements data such as 10 days of temperature, precipitation, sunshine duration, evaporation and others had been spatialy interpolated to get province-wide surface data divided by lkmxlkm space units with higher precision. Mean errors, mean standardized errors and root-mean-standardized errors in 36 10-days precipitation interpolation results are -0.007mm,-0.015 and 1.011 respectively, indicating that the interpolation method used for the interpolation of climate elements is an appropriate and credible ones. And then, climate elements data with the time-series by month and year had been work out through the map algebra methods (raster overlay).2. Result in spatial database establishment of climate elements. Arranging the interpolated climate elements and their relative data, supported by ArcGIS Geodatabase model, with the ArcCatalog software, a spatial database integrating raster datasets and vector datasets including climate elements such as 10-days air temperature, precipitation, accumulated temperature, wetness indices, ground temperature, evaporation amount calculated by 10 days, month and year, land-use areas, administrative divisions, weather stations had been established.3. Results in spatial and temporal variation analysis of climate elements. Through the spatial and temporal variation analysis of climate elements, the spatial and temporal variation laws of climate elements all over the province had been found being as follows.1) The 10-days average air temperature fluctuated over the past 10 years. From the space point of view, the air temperature changes in south-eastern and eastern areas were smaller, but larger in other regions; From the time point of view, the average air temperature during 11-13,12-14 and 19-21 ten days appeared steadily declining trend and steadily rising trend during 15-17,26-28 and 27-29 ten days; The average air temperature is a slight downward trend in July, a slight upward trend in September and November, a rise of about 1.0℃from May to September, greater rising range in plain than hilly areas. The moving average change of annual average air temperature in weather stations fluctuated between 2.5℃and 3.3℃being a mean of 2.9℃appearing a slight rise and fluctuation but no significant upward trend, no significant rising trend compared with the average air temperature for many years. The change ratios of some 10 days of average air temperature during the alternate period from spring to summer reduced to be stable but rised during the turn period of summer and autumn, indicating that there was backward delay in the actual phenological signs.2) 10 days of precipitation and their fluctuations appeared the distribution characteristics that were larger values in the middle period and smaller ones in both ends; So did the values of monthly precipitation but their flucuation degree reversed; Annual precipitation in the area of Xiaoxingan Mountain and Zhagnguangcai Mountain was the largest to reducing to surrounding area gradually, decreasing fastest toward the direction to Daqing City and Qiqihar City and slowly to other directions. The greater peak of regional precipitation for many years reduced, while precipitation differentials among regions decreasing. The value of annual precipitation all over the province interpolated and calculated was 524.6mm, reducing about 60mm compared with historical data. Precipitation in the areas of Xiaoxing’an Mountain and Zhangguangcai Mountain was relatively larger, but smaller in the western of Songnen Plain and the western slope of Daxingan Mountain, being uneven distribution among regions.3) Fluctuations of wetness indices during 13-27 ten days varied greatly in the plain areas, but smaller in hilly areas. The largest variation was in Qiqihar City, while the smallest ones being in Mudanjiang City. The minimum variation was in July, while the maximum being in May in the spring.4) The variations of accumulated temperature in 13-27 ten days in the mountainous and hilly areas and Songnen Plain were the largest, but smaller in Sanjiang Plain. There was an obvious differences between accumulated temperature in Songnen Plain and ones in Sanjiang Plain. Fluctuation variations of 10-days accumulated temperature in Daxinganling Region, Yichun City and Qiqihar City varied greatly, but less in Qitaihe City and Jiamusi City. Annual accumulated temperature of regions and cities in hilly areas was relatively low, higher in the plains and south-east. Fluctuation variations of≥0℃accumulated temperature in Qitaihe City, Jixi City, Jiamusi City and Shuangyashan City varied less, but greatly in Daxinganling Region, Yichun City, Qiqihar City and other cities; Fluctuation variations of≥10℃accumulated temperature in Qitaihe City, Jiamusi City and Jixi City varied less, but greatly in Daxinganling Region, Yichun City, Qiqihar City, Suihua City and other cities.Innovation points:1) This paper carrys on the spatial interpolation of climate elements in Heilongjiang Province by Cokriging Method for the first time, establishing the statistical relationship between climate elements and elevation, aspect, forest cover ratio to interpolate precipitation, evaporation amount, sunshine duration. The spatial varition laws of air temperature, precipitation, accumulated temperature and wetness indices have been worked out based on spatial analysis of ArcGIS software. improving de Martonne model for wetness indices calculation to calculate 10-days wetness indices, making up the defect in previous studies to 10-days climate elements by 10-days scale interpolation, arriving at the regional differentiation laws of climate elements such as air temperature, precipitation, accumulated temperature and wetness indices by spatial analysis with GIS-related software.2) In the course of calculating 10-days wetness indices, the author studied many kinds of wetness index models, considering the de Martonne Model that calculate monthly wetness indices, absorbing the idea that reflects wetness degree through the relationship between precipitation and temperature in this model, to establish a model to calculate 10-days wetness indices in Heilongjiang Province. Calculation of 10-days wetness indices have abtained good effect with the newly created model.3) The spatial interpolation of climate elements carried on mainly by 10-days climate data, acquiring much spatial data of 10-days average air temperature, precipitation, sunshine duration, evaporation amount that have closer relationship to agricultural production, and calculating the monthly and yearly climate data, and the geodatabase of climate data was established based on ArcGIS database technology to lay a foundation to study the physical crop production potential, also made up for the deficiency of spatial distribution data of 10-days climate elements in the previous research.更多还原