香溪河流域土地利用变化过程对非点源氮磷输出的影响

选取三峡库区香溪河流域(3150 km~2),根据1990、2000、2010年Landsat TM/ETM遥感影像,在Arc GIS支持下,应用景观特征分析和氮磷输出系数模型方法,在分析流域景观格局转变过程的基础上,研究了土地利用变化对非点源氮磷输出的影响。结果表明:1990—2000年土地利用变化较为缓和,2000—2010年土地利用转变较为剧烈,变化面积占到总面积的4.3%,为前10年变化量占比的3倍之多。1990—2000年和2000—2010年,土地利用转变量最为明显的均为林地转旱地和旱地转林地;从单位时间土地利用变化率来看,1990—2000年旱地变化最为剧烈,2000—2010年居民地变化最为剧烈;从土地利用相对动态度来看,1990—2000年和2000—2010年旱地均最高。1990—2000年土地利用的变化对非点源总氮(TN)和总磷(TP)输出影响较小,TN增加1.57 t·a~(-1),TP减少0.073 t·a~(-1);2000—2010年土地利用的变化显著降低了TN、TP的输出量,净值分别为78.5、6.1 t·a~(-1)。土地利用转变方式对TN、TP的负荷影响不同,旱地转变为林地时TN负荷表现为消减,林地转变为旱地时TN负荷增加;水田转变为林地时TP负荷减少,旱地转变为水田时TP负荷增加。在输出系数与各土地利用类型面积关系的建立中,使用土地利用状态量变化面积不能真实计算出其对非点源氮磷负荷输出量的影响,通过土地利用过程量面积的变增才能真实反映土地利用变化导致的非点源氮磷负荷输出量。

Effect of land use/cover changes on nitrogen and phosphorus losses via non-point source pathway in Xiangxi River Basin

Land use changes greatly influence non-point source nitrogen and phosphorus outputs. Here, the effect of land use/cover changes on non-point source outputs of nitrogen and phosphorus was studied based on remote sensing images(Landsat TM/ETM) in 1990,2000 and 2010 using the methods of characteristics of landscape analysis, export coefficient model, and ArcGIS in the Xiangxi River basin(3150 km²). Land use/cover changes were relatively moderate from 1990 to 2000, but was severe from 2000 to 2010. The area that land uses had changed from one into another between 2000 to 2010 accounted for 4.3% of the total area, which was three times that from 1990 to 2000. During 1990 to 2000 and 2000 to 2010, the greatest land use/cover type changes were from forest land to dry land or from dry land to forest land. The annual change rates of land use/cover was drastic for dryland from 1990 to 2000 while residential land from 2000 to 2010. The biggest dynamic change area of land use/cover changes was dryland for both 1990 to 2000 and 2000 to 2010. During 1990 to 2000, the land use/cover changes had little effect on total nitrogen(TN) and total phosphorus(TP) loads from non-point source. During 2000 to 2010, however, land use/cover changes significantly reduced the loads of TN and TP, which was 78.5 and 6.1 t·a^-1 respectively. The contribution of different land use/cover changes to TN and TP loads varied. The reduction of TN load mainly happened when dry land changed to forestland. The conversion from paddy field to forestland decreased TP load, whereas turning dryland to paddy field increased TP load. In the establishment of the quantitative relationship between export coefficients and different land use/cover type area, the land use/cover changes calculated by static area could not reflect its impact on the TN and TP loads, while the land use/cover changes calculated by process area was a good indicator of TN and TP loads.