漓江上游水源林地区区域蒸散量的变化规律及影响因素

Lijiang River, best known for world-famous landscape gallery, is the national scenery in China. Recently the water conservation ecological function of forest in the upper reaches of Lijiang River has been degenerating since the water source forest was devastated and the coverage area of broadleaf forest reduced. Although the economic forests such as fur, pine and bamboo were replanted on large area, the phenomenon that low water period of Lijiang River still becomes longer year by year and water flow reduced has not been improved. Forest evapotranspiration is the major factor affecting the hydrological effect and water equilibrium of forest ecosystem. Remote sensing model provides an effective method to study forest ecological function of forest on large area. Therefore, considering the complex topography, the improved SEBAL model was got for southern complex topography basing on the ground measurements, such as meteorological data, surface temperature and evapotranspiration. Then evapotranspiration was retrieved with the improved SEBAL model from the different temporal TM images of the upper reaches of the Lijiang River and validated with the ground measurements. Daily, monthly and yearly evapotranspirations were calculated from instantaneous evapotranspiration with the temporal scaling methods. With the quantitative analysis of evapotranspiration variation and its affecting factors, the effect of forest change on variation of evapotranspiration is evaluated.

漓江以闻名世界的百里"山水画廊"成为国家级风景名胜区。近年来由于上游水源林遭受破坏，阔叶林面积减少，导致森林涵养水源的生态功能降低，尽管大量种植了杉木、松木和毛竹等经济林，但漓江枯水期出现逐年延长、水量变小的现象未见明显改善。森林蒸散是决定森林生态系统水文效应和水量平衡最重要的因素，遥感模型反演蒸散提供了研究大范围地区森林生态功能的一种有效手段。因此，在对漓江上游水源林地区气象数据、下垫面温度、蒸散等实验观测基础上，充分考虑复杂地形的影响，改进SEBAL遥感模型使其适用于南方复杂地形，然后结合TM遥感图像反演不同时相漓江上游水源林区域蒸散量，并用蒸散的观测值验证对应像元的遥感模型估算值，由瞬时蒸散进行时间尺度扩展推算日蒸散量、月蒸散量、年蒸散量，定量研究漓江上游水源林地区区域蒸散量的变化规律及影响因素，分析水源林植被变化对区域蒸散量的影响。

针对上世纪国家级风景名胜区漓江上游水源林阔叶林面积减少，代之以针叶林、竹林等人工经济林导致涵养水源功能降低的问题，为了定量评估漓江上游水源林地区森林植被的恢复状况，考虑复杂地形高程、坡度、坡向的影响，在对漓江上游水源林森林调查、叶面积指数、气象数据、蒸散、下垫面温度等实验观测数据的基础上，通过修正SEBAL遥感模型参数而建立适用模型，利用SEBAL模型和1989–2013年秋季8 景Landsat卫星遥感数据估算森林植被日蒸散量的空间分布，提出了归一化日蒸散量使不同年度秋季日蒸散量具有可比性，分析日蒸散量的时空变化规律。结果表明，利用复杂地形最优多植被指数组合神经网络模型提高了山区林地LAI的估算精度；由于漓江上游森林植被覆盖较高，研究区的日蒸散量空间分布呈现整体数值较高，尤其是漓江水源涵养林猫儿山国家森林保护区保存较好的原始森林植被为日蒸散量的高值区。1989–2013年秋季不同林类的日蒸散量均值，以阔叶林最高，数值在5.00～6.50 mm/d之间，次之为竹林，数值在4.90～6.20 mm/d之间，针叶林的日蒸散量均值在4.20～5.00 mm/d之间，远低于阔叶林；区域日蒸散量除受太阳辐射、气温、风速、相对湿度等气象因素的影响外，林类覆盖变化也是影响研究区日蒸散量的重要因素之一。1989–2000年归一化日蒸散量低值区（数值0.0～0.4）的面积比例由5.8%逐渐上升到14.8%，而高值区（数值0.6～1.0）的面积比例由47.7%下降至32.5%；2000–2013年归一化日蒸散量低值区的面积比例逐渐降低到2009年后均低于10%，而高值区面积比例逐渐上升，到2005年后维持略高于40%，表明研究区森林植被尤其是部分阔叶林逐渐得到恢复。研究结果为遥感监测漓江水源林植被恢复提供科学数据。