Spatial Optimal Placement Of BMPs Based On Nonpoint Source Pollution Distribution Characteristics In Chaohe River Basin

Nonpoint source pollution (NPS) problem can be controlled by implementing various best management practices (BMPs) in the watershed. Although application of conservation practices can effectively reduce nutrients and sediment from plots and fields, simple placement of these mitigation measures in watershed usually can not meet the pollution control target. The reducing efficiency may be influnced by many factors, such as soils, topography, land use, and human activities, etc. Thus, how to optimize spatial placement of BMPs in watershed scale for effective control of NPS with limited funds and labor has been more and more attention from researchers and environmental protection departments. This dissertation took the Chaohe River watershed as the study area and adopted Geographic Information System (GIS), Hydrological Simulation Program-Fortran (HSPF) and Statistical Technique to achieve the objectives which are:1) to identify the spatial distribution frequency of phosphorus pollution loads at two scales (subwatershed and field) and provide the basic data for the development of the conceptual watershed,2) to examine the relationship between probability of statistically significant water quality improvement (Pw) and reduction proportion of pollution load in watershed (Rw)and generate more effect evaluation indices,3) to discriminate the change treands of the four approaches benefit/cost curves for geographically allocating conservation effort. This work will be helpful to further controlling soil erosion and non-point source pollution and protecting drinking water resource watershed. The main results of this doctoral dissertation are listed as follows:1) Based on the spatial distribution features of agricultural non-point source pollution that were drived from simulated experiments. A simple and easy integrated approach will be established for the purpose of identifying the distribution pattern of phosphorus loading at two spatial scales (watershed and filed) so that different allocation plans of BMPs can be designed. More effective performance benefit index will be generated through identifying relationship between the ratio of TP loading mitigation and the probability of detecting a statistically TP improvement. By comparing the thresholds of logistic curve of cost-benefit, the best allocation plan will be determined. The expected results will be very useful for agricultural nonpoint source pollution control and drinking water source protection in Chaohe River watershed.2) The HSPF model simulation results can meet the need of conceptual subwatershed construction. The phosphorus load from Chaohe River flowed into Miyun reservoir mainly in June-September (flood season). The loss of sediment in flood season accounted for78%-90%of the whole year load. Thus, the flood season is a critical period for soil erosion control and non-point source pollution prevention. For each subwatershed, the average pollution load was21.32kg, the max value was113.85kg and the min value was0.91kg, respectively. The phosphorus index (PI) results showed that frequency distribution of field P load estimates was best fit by a normal probability distribution. High risk area, moderate risk area and low risk area account for7.95%,19.63%,72.42%of the total area. PI value has a significant correlation (R2=0.67) with the actual loss value at watershed scale. The average values of phosphorus from field scale is2.94kg, the max and min value were17.24kg and2.07kg, fitting a spatial normal distribution. Therefore, it can be used as data basis for the conceptual watershed construction.3) Conceptual watershed included100model subwatersheds and10000model fields, which were selected for random order among the Chaohe river basin. The pollution load of model landscape had spatial normal probability distribution feature. The phosphorus concentration had significant spatial and temporal difference. The dataset consists of time series of P concentrations from thirty tributaries with more than40%watershed agriculture and provides the best available estimate of P variability in Chaohe River. Based on statistical simulations and stream P concentrations data, a statistically significant waterquality improvement at the outlet of a watershed (PW) was calculated as:Pw=1.059/1+69(e-13.2·Rw)-1.038In realistic implementation scenarios, the aggregated and targeted approach most efficiently improves water quality. The sensitivity analysis (reduction proportion of fied load, load spatial distribution and sample number) indicated that the superiority of the aggregated and targeted approach was robust to uncertainties in model parameters and to assumption choices.