Analysis of snowmelt model for flood forecast for water in arid zone: case of Tarim River in Northwest China

• Published in: Environmental earth sciences Vol. 66; no. 5; pp. 1423 - 1429
• Main Authors: de la Paix, Mupenzi Jean, Lanhai, Li, Jiwen, Ge, de Dieu, Habumugisha Jean, Theoneste, Nzayisenga
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer-Verlag 01.07.2012; Springer Nature B.V
• Subjects: accuracy; Arid zones; Biogeosciences; Climatic conditions; climatic factors; computer software; Earth and Environmental Science; Earth Sciences; Environmental science; Environmental Science and Engineering; Flood forecasting; Floods; Freshwater; Geochemistry; Geology; hydrologic models; Hydrology/Water Resources; mathematical models; Original Article; prediction; River basins; Rivers; Runoff; Snow cover; Snowmelt; snowpack; Sublimation; temperature; Terrestrial Pollution; Water availability; watersheds; China; Tarim Basin; China, People's Rep; D-IUH runoff model; D-IUH model estimation; Snowmelt model; ARIMA model; Flood forecast
• ISSN: 1866-6280; 1866-6299
• DOI: 10.1007/s12665-011-1353-9

Field studies were carried out in Tarim River Basin, Northwest China for analysis of snowmelt model for flood forecast for a river in arid zone. Snow is a major source for water availability in arid zone of Northwest China where 50% of snow cover withdrew by sublimation during dry and cold climatic condition. The analysis of weekly forecast of daily discharges was helped by the temperature index model, ARIMA model for temperature and flow, D-IUH runoff model and D-IUH model estimation where the temperature forecast was used as driving variable; the numerical simulations were carried out using SUSA® software for testing the sensitivity of the D-IUH to the input values of the parameter and an analysis of the forecast results against the set of input parameters resulted in a determination coefficient R 2 = 0.5. The standard deviation was 3.28 and the mean for the Tarim River was 5.37 (mm d−1) implying that the forecasted data is in strong agreement with the observed data. The combination of methods is better useful for calculation in order to avoid errors of appreciation.