Basin Hydro-cycle And Associated Non-point Source Simulation Research And Application Based On Distributed Hydrological Model

Water is an essential natural resource for human life and social activity. It is also a most important environmental and domestic production resource for dependents. As a natural matter, water moves in cycles driven by the atmospheric stress and hydrological processes such as evaporation, precipitation, and runoff. It supports the vital and physical phenomena. It shares both natural and social functionality. On one hand, water is a kind of solvent that major type of substances can be dissolved in water. The other is that water has potential, kinetic, chemical energy supporting flow and dissolution. Besides the natural functionality, water also has unique social function that it is an indispensable element for human society. Besides the heterogeneity in water quantity and spatial distribution that may affect human activity. Water quality is another significant factor. With ever increasing speed of industrialization and urbanization, the world is challenged by water resource shortage and water pollution. Till now, water pollution is mainly caused by non-point source pollution including soil loss, rural living, fertilization and livestock raising. The non-point source pollutant increases the nutrients in water body, decreases the oxygen concentration thus causing eutrophication. destroying water ecology, affecting human activity, even threatening human healthy. Since the beginning of 21th century, distributed hydrological model (DHM) has become a hot spot in watershed hydro-cycle research. Computer technology, geographical information system and remote sensing tech. have been widely used in distributed hydrological research. With the combination of these technologies, DHM can better describe the underlying and meteorological heterogeneity.Therefore, based on the dualistic "natural-social" theory, the study puts forward a distributed hydro-cycle and associated process simulation scheme, by developing DHM, Sensitivity analysis and calibration method, hydro-cycle based distributed non-point source pollution model. It is meaningful to study on the watershed hydro-cycle and pollutant transfer, water quantity and quality analysis.The outline and research results are listed:(1) Research on basin distributed hydro-cycle and associated non-point source process schemeThe scheme of distributed hydro-cycle simulation and associated non-point source simulation, high dimensional model parameter optimization and integration tech. was put forward. EasyDWAT (Easy Distributed WaterResource Assessment Tool), SenOpt (Sensitivity with Optimization), EasySNP (Easy Statistical Non-point source Pollution assessment) and EasyDNP (Easy Distribted Non-piont source Pollution assesssment) were put forward. The integration between SenOpt and EasyDHM, EasyDHM and EasyDNP, EasyHydro and EFDC (Environmental Fluid Dynamic Code) were given.(2) The development of hydro-cycle simulation model and non-point source modelThe development process and method on distributed hydro-cycle and associated non-point source simulation model were describted. Monthly distribted hydrological model EasyDWAT based on EasyHydro and water resource trend analysis method were given. The water balance equation and underlying parameters were discussed. Detailed description of EasySNP and EasyDNP were given. These models were integered into EasyHydro in order to improve each other.(3) Sensitivity analysis and multi-objective optimization of DHMHigh model parameter dimension brings low convergences in model optimization. In order to solve this problem, the study put forward a joint sensitivity and multi-objective optimization method called SenOpt. The study took the Zhuba catchment in upstream Yangtze River as an example. The analysis began with the global sensitivity of EasyDHM, and then parameter selection is conducted according to the parameter ranking and sensitivity level. Results showed that the parameter convergence speed is much higher by taking sensitive parameter than embracing all parameter in optimization. Meanwhile, the study adopted MPI distributed evaluation scheme to modify the parallel complex evolution, saving computational load in optimization.(4) Distributed hydro-cycle simulation and water resource change study using EasyDWAT mdoelThe study analyzed meteorological and hydrological factors in upstream Yangtze River. The study obtained the trend and cycle of precipitation, air temperature and runoff. Then, the model structure of EasyDWAT was detailed explained. The model took raster as the computational unit, take full consideration of underlying landuse and soil type. Taking the upstream Yangtze River as a study case, the study developed a monthly EasyDWAT model using input material from year 1956 to 2005 and drawing considerable simulation results. Meanwhile, the model also outputed respectively the spatial distribution of seasonal and yearly averaged precipitation, evapotranspiration, surface runoff and groundwater recharge. Detailed explanation on the water resource spatial distribution of upstream Yangtze River corresponding to the variation of landuse was also given, drawing ideal validation.(5) Hydro-cycle based point and non-point source load evaluationThe study described the structure of distributed non-point source statistical model EasySNP and distributed non-point source simulation model EasyDNP. The study took the Three Gorges area as a study case and developed a statistical model considering hydrological boundary and political boundary. The model took point source and non-point source into consideration. The point source included industrial and urban point source while non-point source includes rural living, livestock rising and fertilization. Meanwhile, based on distributed hydrological model EasyDHM, the study also used EasyDNP model to evaluate the load of Three Gorges area. The study results of EasyDNP and EasySNP were compared and analyzed. Results showed that load computed using EasySNP based on hydrological boundary was lower that political boundary. And the load simulated using EasyDNP model was lower than EasySNP. The hdyro-cycle based EasyDNP was able to take more source-sink processes of non-point pollution, the simulation results was closer to the reality. Further, the study also discussed the changing process under different precipitation scenarios.(6) Integration of distributed hydrological model and EFDC with applicationTaking upstream catchment of Miyun reservoir as study case, the study develops DHM model EasyDHM by integrating EasyDNP to simulate monthly runoff and pollutant load. The runoff and load time series was used as upstream boundary of Miyun reservoir EFDC model to simulate water temperature, dissolved oxygen, chlorophyll, total nitrogen and total phosphorus. The simulation results showed ideal fitting between simulated water quality item and the observation.