Spatial-temporal Patterns Of Vegetation Evapotranspiration In The Upper Minjiang Valley, China

Evapotranspiration (ET) is an essential component of water balance including evaporation and plant transpiration. The multi-disciplinary knowledge of meteorology, agrology, botany and hydrology is needed to understand ET. The increasing understanding of ET is critical for exploring mechanism of the global water cycle as well as that of ecosystem pattern and hydrological processes. About 60% of precipitation was transported back into atmosphere by ET every year, while the proportion can reach 99% in croplands. ET, as an important component of water cycle, plays a significant role in the research of water cycle process, vegetation regulation to water process and sustainable use of limited water resources.Based on remote sensing and meteorological data as well as field study, the following studies were carried out in the upper Minjiang Valley by the means of 3S. (1) Spatial analysis of vegetation pattern at a landscape scale; (2) Modeling of ET spatial-temporal patterns based on different methods (Thornthwaite and SEBAL model were used); (3) The relationship between ET and vegetation and meteorological factors. The following results were concluded:1) The spatial analysis of vegetation pattern at a landscape scale shows the dominant vegetation are composed of conifer forests, shrub lands and grasslands, Their areas account for 30.26%, 22.76%, 36.88% of the total, respectively. The shrub lands have the biggest patch number and density while they have poor connectivity; Conifer forest and grasslands have high contagion and connectivity. The Wolong catchment has the lowest fragile landscape pattern while Heishui and Zhenjiangguan catchments are most fragile under continual human disturbances.Analysis also shows vegetation distribution is correlated to elevation, solar radiation, temperature, precipitation and so on. This correlation is not significant at the whole study area because of the diverse microclimate and uncertain human disturbances, while the correlation is high at the catchments scale due to relatively even vegetation distribution.2) Based on remote and meteorological data, the ET of the upper Minjiang Valley was modeled by using different methods including Thornthwaite and SEBAL model. Thornthwaite model was modified by a montane solar radiation hour calculation method in order to accurately estimate mountain ET. The prediction error by modified Thornthwaite is 5% lower than unadjusted model, which is not a big deal but an important result. The result shows that the solar radiation hour is an important factor influencing ET especially in montane area.SEBAL (Surface Energy Balance Algorithm for Land) is a popular mode! developed by Bastiaanssen (1995) as a remote sensing data based model that derives ET as the residual of surface energy balance. The model has been applied in various ecosystems in many places of the world. The instantaneous ET of the study area was simulated by SEBAL; and the net...