| Based on the multidisciplinary theoretical system such as ecology,climatology,hydrology,remote sensing and geographic information system,the goals of this paper were to develop a suitable methodology to estimate ecological water requirement of vegetation in humid regions of South China and study on its spatiotemporal variations in the context of climate change by making comprehensive use of computer software programming,remote sensing information extraction technology and geospatial analysis,etc.Taking Hunan Province as a case study,the paper selects crop water requirement and the optimum water requirement of woodland as research indexes and realizes the calculation of these indexes under the support of RS and GIS technologies.From the perspective of geospatial information,the paper draw some conclusions on spatiotemporal variations of climatic variables,reference evapotranspiration,vegetation ecological water requirement and ecological water shortage etc.over large spatial and temporal scales.The ecological water requirement calculation system of Hunan Province was also established.The research results of the paper can further promote the study and application of mapping vegetation ecological water demand as well as its relevant indexes over a large spatial scale and long-time scale in humid regions of South China.Also,they are conducive to the rational utilization and optimal allocation of water resources,agricultural development and forest protection,etc.for the study area.Moreover,they can provide a reference for relevant management and decision-making departments in the hope of a conversion from purely natural ecological water consumption gradually to artificial rational deployment of high-efficient water-saving consumption.The results showed as follows.(1)The paper integrated multi-temporal remotely sensed images with several key land surface phenological parameters and used an object-oriented Random Forests classifier by applying multi-resolution segmentation method to these remotely sensed data in order to map continuous annual land use and land cover across Hunan Province during 2000-2015.The algorithm is robust and the error matrix indicated that classification accuracies of all types of land use and land cover are basically accepted.Also,the training dataset,validated by Landsat ETM+/OLI and Google earth high resolution images and phenology,can help to improve classification accuracy.The results showed that the spatial patterns of land use and land cover in Hunan province is relatively stable over time,but differed markedly in the areal coverage,and that transfer matrix demonstrated mutual transformation among all land use and land cover types was quite drastic.(2)The paper calculated ET0 by using the Penman-Monteith equation and data from 87 meteorological stations in Hunan Province from 1960 to 2015.Results indicated that the declining trend of annual ET0 is mainly explained by the decline in sunshine duration and wind speed.On the contrary,dropping in relative humidity increases annual ET0 to some extent.ET0 in spring was found to increase significantly attributing to the decrease in relative humidity as well as the rise in maximum temperature,while the downward trend of ET0 in summer,autumn,and winter was mainly related to the decline in sunshine durition and wind speed.Annual wind speed,relative humidity and sunshine duration decreased while temperature,precipitation and wetness index showed an upward trend,suggesting a warming-wetting trend in Hunan Province.Spatial distribution of ET0 in Hunan Province under changing climate differed significantly among various regions both over annual and seasonal scale.(3)In this paper,the daily precipitation data of 87 meteorological stations in Hunan Province from 1960 to 2015 are used to calculate effective precipitation during crop growth periods and during the growing seasons of different woodland vegetation.Generally,effective precipitation of the early season rice showed an upward trend,while the late season rice,closed forest land,sparse wood land,shrub land and other forest land indicated a declining trend.Geographically,annual effective precipitation of the early season rice were more abundant in the southern and western regions of the study area.While,effective precipitation of the late season rice presented a decreasing trend from the southeast and northwest to the central areas inside the study area.Also,the spatial pattern of effective precipitation of those woodland vegetation was more complicated with no obvious regularity under the influence of topography.(4)The paper obtained evapotranspiration during crop growth periods and during the growing seasons of different woodland vegetation through multiplying ET0 by various vegetation coefficients.During 1960-2015,evapotranspiration of the early season rice,late season rice as well as those woodland vegetation revealed a downward trend,among which evapotranspiration of the late season rice decreased significantly.Geographically,evapotranspiration of the early season rice in the eastern areas are relatively stronger than the western parts of the study area,while evapotranspiration of the early season rice presented an increasing trend from the northwest to the southeast inside the study area.Meanwhile,evapotranspiration of those woodland vegetation mainly showed an upward trend from the northwest to the southeast inside the study area with obvious strip distribution.(5)The paper comprehensively and systematically analyzes the spatiotemporal variations of crop water requirement,ecological water requirement of different woodland vegetation and total vegetation ecological water requirement in the region.During 1960-2015,crop water requirement,woodland ecological water requirement and total vegetation ecological water demand all showed a downward trend,among which crop water requirement decreased significantly.Geographically,the distribution of crop water requirement were more abundant in the northern and eastern regions of the study area,and the water requirement in the plains,valleys and basins is relatively high.On the one hand,the spatial pattern of various woodland vegetation ecological water requirements is mainly affected by topography,for example,the water requirements in mountainous and hilly areas are relatively high.On the other hand,they also depends on the distribution range of different types of woodland vegetation,the larger the vegetation distribution area,the greater the ecological water requirement,and conversely,the smaller the distribution range,the smaller the water requirement.(6)The paper comprehensively and systematically analyzes the spatiotemporal variations of irrigation water requirement,crop water shortage,ecological water shortage of different woodland vegetation and total vegetation ecological water shortage in the region.During 1960-2015,irrigation water requirement,crop water shortage,and total vegetation ecological water shortage all showed a downward trend,while each woodland vegetation ecological water shortage indicted an upward trend.Geographically,the spatial patterns of irrigation water requirement of the early season rice and crop water shortage were basically consistent with the pattern of crop water requirement,while the spatial pattern of irrigation water requirement of the late season rice mainly presented an increasing trend from the southeast and northwest to the central areas inside the study area.The spatial pattern of each woodland vegetation ecological water shortage depends on corresponding ecological water requirement and effective rainfall during the growing season.Generally,with a wide range of certain type of woodland vegetation,the water requirement is large and the shortage is also high.But because of the rise in effective precipitation due to orographical rain may decrease woodland vegetation ecological water shortage.(7)Based on the VS2010-C#-Winform programming tools,iGeoVEWRs v1.0 was established to create program scripts to quickly calculate crop evapotranspiration,effective precipitation,water requirement,water shortage and artificial rainfall. |
PDF Full Text Request