【作者】 马玉平；

【导师】 李维京；

【作者基本信息】 中国气象科学研究院 ， 气象学， 2011， 博士

【摘要】 植被覆盖状况的改变不仅直接影响陆地生态系统的形成和演变,而且还可能引起不同尺度的大气环流和气候变化。同时,气候变化也直接影响陆面植被状况的改变。因此,植被大气相互作用及其反馈机制已成为当前全球变化研究中的重要科学问题之一。我国西部干旱半干旱地区气候条件恶劣、生态环境脆弱且经济落后。1999年以来,该区逐步实施了退耕还林还草生态建设工程。这一土地利用方式的巨大转变可能引起当地气候的变化,并进而影响农业产生。如何客观评估退耕还林还草的气候和农业效应,是涉及当地社会经济和生态环境可持续发展的重大问题。本研究以陕甘宁降水较少,林木立地条件差,退耕还林还草为主的地区为例,首先根据观测数据分析当地植被变化与区域气候间的相互影响关系,然后探讨区域气候模式与作物/牧草生长模式的耦合方法及耦合效果,最后结合退耕还林还草的实际进展情况,利用耦合模式长年代多样本的数值模拟试验,重点研究土地利用渐变条件下的气候-植被相互作用,探讨单纯由退耕还林还草所产生的气候和农业效应。主要研究结论有:1.观测研究表明,自2000年退耕还林还草以来陕甘宁地区的主要土地利用类型分布发生了显著性的变化,冬小麦种植面积由于退耕而每年缩小约1568 km²,草地面积由于转换为半沙漠、冬小麦种植区或林地等原因而每年减少2624 km²,沙漠面积由于防沙治沙工程的实施而每年减少1776 km²,落叶阔叶林由于植树造林而每年增加3312 km²,半沙漠由于草地退化、治沙、退耕地还未发展成草地或林地而每年增加2300 km²;与研究区北部非退耕区相比,中南部典型退耕区夏季高温随时间的增速趋缓而降水量随时间的增速变快;夏季温度变化与落叶阔叶林、半沙漠、沙漠面积的变化显著相关;春季温度和降水量对冬小麦产量有明显影响,而秋季温度和年降水量影响着牧草产草量。这些观测事实是开展退耕还林还草气候和农业效应模拟研究的前提。2.模式校准和性能分析表明,作物生长模型WOFOST中加入干物质再运转过程提高了模型在单点上的模拟性能,对冬小麦参数的分区实现了模型的区域应用。牧草生长模型LINGRA中加入茎秆生长过程和钝化水分胁迫系数后能够较好地适应对天然牧草的模拟。网格（水平分辨率为20km）和次网格（水平分辨率为4km）尺度下的模拟检验表明,区域气候模式RegCM3能够合理再现陕甘宁地区气候的季节变化和区域分布特征,但模拟温度稍偏低,模拟降水量偏高。三个模式的校准和性能分析为应用研究奠定了基础。3.模式耦合设计和试验结果表明,RegCM3与RegWOFOST\LINGRA可以在日和次网格尺度上实现时空匹配。RegCM3向RegWOFOST\LINGRA输出最高、最低温度、降水量、辐射、水汽压和风速等气象数据,RegWOFOST\LINGRA以叶干重与比叶面积之积的方法计算植被LAI,再反馈回RegCM3而实现两类模式的耦合。耦合模式RegCM3\WOFOST\LINGRA在一定程度上改善了RegCM3模拟降水量偏高、温度偏低的现象,提升了RegCM3的模拟性能,并实现了气候、作物和牧草的同时模拟,为气候和农业效应研究备好了工具。4.退耕过程中（1999-2008年）RegCM3\WOFOST\LINGRA的数值模拟结果表明,单纯的退耕还林还草导致研究区多数地区最高、最低和平均温度明显下降,太阳总辐射略有下降,而降水量、蒸散以及土壤湿度增加的区域均较减小的范围更广,数量更大,地表径流降水比变化较小,增减范围大小相当。从时间变化看,单纯的退耕还林还草对研究区范围的日最高温度持续升高有一定的抑制趋势,对南部典型退耕区的抑制趋势更明显,对降水量的增加也有微弱促进作用,并由此导致冬小麦成熟期延迟,贮存器官干重等产量要素明显增加。其中的降温主要与耕地或草地转换为林地相对应。5.退耕前后RegCM3\WOFOST\LINGRA的长期（1990-2015年）数值模拟表明,单纯的退耕还林还草造成陕甘宁全区域平均的最高温度有所降低,降幅在0.05-0.27℃之间,最低温度降低在0.1-0.3℃之间,降水量增加在25mm以内,太阳总辐射减少在80mj m^-2以内,由此导致冬小麦成熟期提前2天左右,开花期LAI增加在0.2以下,成熟期贮存器官干重增加在614 kg hm^-2以下。其中,平均气温和降水量的变化占总的气候变化的比重在3%～7%之间,对其它气候和农业要素的影响则更大一些。表明目前的退耕还林还草所产生的气候效应是探讨区域范围内气候变化不可忽略的部分,但它尚不能逆转全球气候变化带来的影响。本研究在揭示观测事实、开展区域气候模式与作物/牧草生长模型耦合、量化单纯有退耕还林还草引起的气候和农业效应、考虑实际土地利用变化状况、以及在较高空间分辨率进行长时间数值积分和多样本模拟试验等方面具有一定创新。本研究针对气候与植被之间相互作用及其反馈机制的科学问题,在研究的方法论上进行了有益的尝试。本研究既促进气候和农业学科交叉研究的发展,同时又具有一定社会效益。下一步还可在模式性能改善、耦合的切入点、更大研究区域、空间气候变量效应分析等方面继续开展研究。更多还原

【Abstract】 Changes of vegetation cover not only directly affect the formation and evolution of terrestrial ecosystems, but also may give rise to different scales of atmospheric circulation and climate change. At the same time, climate change has a direct impact on changes of land surface vegetation. Thus the interaction and feedback mechanism between vegetation and atmosphere becomes one of the important scientific issues in current global change research. There are poor climatic conditions, fragile ecological environment and backwardness economic in arid and semiarid climate regions in West China. The Ecological Reforestation Project had been implemented since 1999 in there. This land use change is bound to cause great changes of local climate and to affect agriculture. How to assess the possibility climatic and agricultural effects of returning farmland to forest and grassland is the major issue involved in local sustainable development of socio-economic and ecological environment.In the study, based on the region with less precipitation, poor forest site, converting farmland into forest or grass land in ShanGanNing, the relationship between local vegetation change and regional climate was firstly analyzed by using some observations. Then, it was investigated that coupling method and its effect between regional climate model and crop / pasture growth model. Finally, interactions between vegetation and climate were researched under the conditions of gradient changes of land use in the actual progress. Climate and agricultural effects produced by Alone Returning Farmland to Forest and Grassland （AReFFG） were discussed through many numerical simulations by using the coupling model. The main conclusions were as follows:1. Analysis of observational data showed that （1） Distribution of major land use types in ShanGanNing region had produced significant changes since 2000. Planting area of winter wheat reduced by about 1568 km² y^-1 due to returning farmland to forest or grassland. As a result of degradation or converting to farm land or forest land, the area of grass land reduced by 1776 km² y^-1. There were reductions of 1776 km² y^-1 in desert area since the implementation of desertification control projects. Deciduous broad-leaved forest area increased by 3312 km² y^-1 based on afforestation. Because of conversion from farmland or degradation of grassland, semi-desert / sparse grass area increased by 2300 km² y^-1. （2） Increasment slowed down in July temperature while accelerated in precipitation with time in Typical region of Ecological Reforestation in Central and Southern study Area （TERCS region）. （3） Summer temperature was significantly associated with changes of deciduous broad-leaved forest, semi-desert and desert area. （4） Spring temperature and precipitation had a significant effect on the yields of winter wheat, while autumn temperature and annual precipitation affected the amount of forage grass. These facts from observations were a prerequisite to carry out simulation studies on climate and agricultural effect of converting farmland into forest or grass land.2. Calibration and validation of models showed that （1） the performance of WOFOST in local scale was improved by adding re-operation process of the dry matter. Regional application of WOFOST was achieved based on regionalization of winter wheat parameters. （2） LINGRA could not simulate the growth of natural grass until including stem growth process and making the water stress coefficient insensitive. （3） RegCM3 could reasonably reproduce the seasonal changes and regional distribution of climate in ShanGanNing region under both grid and semi-grid scale. However, there were slightly lower in simulated temperature and higher in simulated precipitation than observations.3. Coupling design and test of models showed that （1） RegCM3 and RegWOFOST\LINGRA could match in days and semi-grid scale. Both models were coupled by means of that RegCM3 outputted the maximum, minimum temperature, precipitation, radiation, vapor pressure and wind speed to RegWOFOST\LINGRA, and vegetation LAI, which was calculated by using leaf dry weight and specific leaf area in the latter, was then fed back the former. （2） RegCM3\WOFOST\LINGRA slightly improved the situation of high in precipitation and low in temperature simulated by RegCM3 and achieved to simulate simultaneously the climate, crops and pasture.4. Numerical simulation by using RegCM3\WOFOST\LINGRA during 1999 to 2008 showed that （1） As a result of AReFFG, there were a decrease in the maximum, minimum and average temperature, a slight decrease in global solar radiation in most parts of the study area, the region of increase in precipitation, evapotranspiration and soil moisture was wider than the one of decrease, and the ratio of surface runoff and precipitation had a little change. （2） AReFFG had inhibitory effect on continuous increase of the maximum temperature in study area and promoted effect on the increase of precipitation in TERCS region. These situations led to noticeable delay in winter wheat maturity and increase of dry weight of storage organ （WSO）. Temperature change was mainly corresponding with conversion of farmland or grassland into forest land.5. Numerical simulation results by using RegCM3\WOFOST\LINGRA during 1990 to 2015 showed that （1） As a result of AReFFG, there were a decrease of 0.05℃-0.27℃in the maximum temperature, a decrease of 0.1-0.3℃in the minimum temperature, a reduction of less than 80mj m^-2 in the global solar radiation, a increase of less than 25mm in precipitation, a advancement of about 2 days in winter wheat maturity, a increase of less than 0.2 in LAI in anthesis, an enhance of less than 614 kg hm^-2 in WSO in maturity in average of study area. （2） The proportion of average temperature and precipitation changes caused by AReFFG were 3%-7% in the total climate changes, while other climate and agricultural elements account for the larger proportion. Climate and agricultural effect of AReFFG could not be neglected in probing regional climate change. However, it still could not reverse the large background of climate change.In this paper, there were innovation in revealing the observed facts, coupling regional climate model and crop / pasture growth model, trying to separate the effects of AReFFG, considering the actual change situation of land use, higher spatial resolution and long-time numerical integration and many numerical simulation test as well. We had a useful attempt in study methodology against scientific issues of the interaction and feedback mechanism between vegetation and climate. This research not only promotes the interdisciplinary study of climate and agricultural, but also has social benefits. It can be further investigated to improve model performance, add more entry point of models coupling, and consider the situation of larger area and effects of space weather variables in the next step.更多还原