亚热带典型农业流域河流水质多元线性回归预测

发布时间：2018-05-14 18:38

本文选题：流域环境 + 氮磷　；参考：《生态与农村环境学报》2017年06期

【摘要】：理解河流水体氮磷浓度与景观格局之间的关系是流域水环境质量科学管理的前提。选取湖南省长沙县石燕河流域,基于Pearson相关性分析、方差分解分析和多元线性回归方法,研究景观格局对流域出口水体氮磷浓度的影响并构建氮磷浓度预测模型。结果表明,流域出口水质长期处于劣V类,铵态氮(NH_4~+-N)、硝态氮(NO3--N)、总氮(TN)、可溶性磷(DP)和总磷(TP)质量浓度分别为5.67~22.46、0.76~2.85、13.41~45.55、0.86~5.00和1.99~9.94 mg·L~(-1);流域出口水体氮磷浓度与土地利用方式组成、景观格局指数、人口和养殖密度相关性显著(P0.05)。流域出口水体TN和TP浓度与林地面积比例、养殖密度呈显著正相关(P0.05),与农田、居民地面积比例和人口密度呈显著负相关(P0.05);TN和TP浓度与景观格局最大斑块指数(LPI)呈显著正相关(P0.05),与景观斑块形状指数(LSI)呈显著负相关(P0.05);流域出口水体氮磷浓度可用土地利用方式组成(农田、居民地、茶园)和景观格局指数[斑块密度(PD)、LPI、LSI、蔓延度(CONTAG)、景观分割度(DIVISION)]的多元线性回归模型预测(校正后R2为0.132~0.320),且多元线性回归模型对NO_3~--N浓度的预测效果最佳。研究结果可为亚热带丘陵区农业流域水环境保护与景观合理规划、利用和管理提供科学依据。
[Abstract]:Understanding the relationship between nitrogen and phosphorus concentration and landscape pattern is the premise of water environmental quality scientific management. Based on Pearson correlation analysis, variance decomposition analysis and multivariate linear regression method, the effects of landscape pattern on nitrogen and phosphorus concentration in water body of Changsha County, Hunan Province were studied and a prediction model of nitrogen and phosphorus concentration was established. The results showed that the quality of export water in the basin was in the category V for a long time. The mass concentrations of NH _ 4- N ~ -N, no _ 3-N _ 3- N _ 3, total N _ N _ N _ N _ N _ 3- N _ 3-N _ 3, total N _ N _ N _ N _ N _ N _ 3- N _ 3- N _ 3- N _ 3) and TP _ (TPN) were 5.67 ~ 22.46 ~ 0.76 ~ 2.86 ~ (0.85) ~ (0.865.00) and 1.999.94 mg / L ~ (-1), respectively, and the composition of nitrogen and phosphorus in the export water of the basin was 1.999.94 mg / L ~ (-1), respectively. Landscape pattern index, population and culture density were significantly correlated with P0.05. The concentration of TN and TP in the outlet water of the watershed was positively correlated with the proportion of forestland area, and the culture density was positively correlated with the farmland. There was a significant negative correlation between land area ratio and population density (P 0.05), TN and TP concentrations were positively correlated with the maximum patch index (LPI) of landscape pattern, and negatively correlated with landscape patch shape index (LSI). Made up of land use (farmland, The multiple linear regression model (R2 = 0.132 ~ 0.320) and landscape pattern index (patch density, landscape division) were used to predict the concentration of NO_3~--N, and the results showed that the multivariate linear regression model had the best effect on the prediction of NO_3~--N concentration. The results can provide scientific basis for water environment protection and landscape rational planning, utilization and management of agricultural watershed in subtropical hilly region.
【作者单位】：中国科学院亚热带农业生态研究所亚热带农业生态过程重点实验室;中国科学院亚热带农业生态研究所长沙农业环境观测研究站;中国科学院大学;湖南省土壤肥料研究所;湖南农业大学资源环境学院;湖南省水利厅;
【基金】：国家科技支撑计划(2014BAD14B01) 城郊环保型高效农业关键技术研究与示范(2014BAD14B05) 湖南省水利科技项目(湘水科计[2016]-SB-01) 湖南省科协决策咨询研究计划
【分类号】：X52;X71

【相似文献】