A Study On Snowmelt Runoff Forecasting And Its Uncertainty Analysis In Xinjiang

Monitoring of snow along with simulation and forecasting of snowmelt has become the research focuses in both climate change science and hydrology. But researches in snowmelt prediction are yet at the initial phase, especially rare in uncertainty studies, reasons of which are widespread uncertainties in snowmelt process, lack of observed experimental conditions, and weak research infrastructure.This paper aims at two cardinal scientific issues - key techniques and methods in distributed snowmelt runoff forecasting and uncertainties in snowmelt hydrological prediction, to systematically and deeply address the prediction of Xinjiang snowmelt floods and their uncertainties. The Juntanghu Basin, which located in the Changji State in the north slope of Tianshan Mountains in Xinjiang, was choosed to be the typical study area, applied "3S" technologies (GIS, RS and GPS), laboratory simulation, field observation experiments, computer modeling and other modern mathematical and technical means, and carried out a large number of field observations and experimental data collections as well as laboratory simulation and analysis of the physical process of the snowmelt.Major contents of studies and conclusions in this paper are as follows:First, basic theories and methods for snowmelt runoff forecasting are discussed in this paper. Second, a series of simulations of snow freezing and thawing process were carried out, to build the snowmelt model and to provide basis for its parameters calibration. Third, it takes research on construction of the distributed snowmelt runoff model, and a Distributed Snowmelt Runoff Model was initially constructed for Xinjiang area. Fourth, researches based on the atmosphere - land surface coupling mode combination with conducting the model parameter calibration and distributed data-input are completed. Fifth, research of the uncertainty focused on the process of snowmelt and snowmelt runoff hydrological model is another important part of this paper. Sixth, measurement standards of prediction uncertainty and parameters calibration of the snowmelt model, as well as reliability analysis, are on the table for systematic discussion. Seventh, it launches research works on the theory and method of the uncertainties assessment of the snowmelt hydrological process and snowmelt models.The research on the snowmelt runoff forecasting and its uncertainty, which is based on the social needs of the regional water resources management and the disaster prevention and mitigation in Xinjiang, is one of the forefront research focus and scientific issues of international hydrology science, in view of this, this study is the first series of international standards, the main innovations include two aspects:First of all, based on the atmosphere - land surface coupling modes, "3S" technologies, and indoor and outdoor simulation experiments and observations, a distributed snowmelt runoff model is set up for Xinjiang. Combining with the outputs of atmospheric modes, multi-source data such as remote sensing data and basic geographic data, were used to drive the distributed snowmelt runoff model and forecast snowmelt runoff, absolutely, this is the direction of the hydrological forecasting in the future.The second part of innovative work is that, it takes embedded and systematic discussion on uncertainty of the snowmelt hydrological processes and snowmelt runoff forecasting, as well as the relevant theories and research methods of them. Systemic research on the uncertainty of snowmelt is the first time and it's among the forefront in the world.