In this study, recurrent networks to downscale meteorological fields of the ERA-40 re-analysis dataset with focus on the meso-scale water balance were investigated. Therefore two types of recurrent neural networks wer...In this study, recurrent networks to downscale meteorological fields of the ERA-40 re-analysis dataset with focus on the meso-scale water balance were investigated. Therefore two types of recurrent neural networks were used. The first approach is a coupling between a recurrent neural network and a distributed watershed model and the second a nonlinear autoregressive with exogenous inputs (NARX) network, which directly predicted the component of the water balance. The approaches were deployed for a meso-scale catchment area in the Free State of Saxony, Germany. The results show that the coupled approach did not perform as well as the NARX network. But the meteorological output of the coupled approach already reaches an adequate quality. However the coupled model generates as input for the watershed model insufficient daily precipitation sums and not enough wet days were predicted. Hence the long-term annual cycle of the water balance could not be preserved with acceptable quality in contrary to the NARX approach. The residual storage change term indicates physical restrictions of the plausibility of the neural networks, whereas the physically based correlations among?the components of the water balance were preserved more accurately by the coupled approach.展开更多
With climate change, water may become limited for intensive agriculture even in regions presently considered “water-rich”. Information about the potential water requirement and its temporal and spatial variability c...With climate change, water may become limited for intensive agriculture even in regions presently considered “water-rich”. Information about the potential water requirement and its temporal and spatial variability can help to develop future water management plans. A case study was carried out for Switzerland with its highly complex pre-alpine topography and steep gradients in climate. The hydrological model WaSiM-ETH was used to simulate net irrigation requirement (NIR) for cropland, grassland and orchards using criteria to define irrigation periods based either on the water stress level (expressed by the ratio of actual (aET) to potential evapotranspiration ((pET) (Method 1) or on thresholds for soil water potential (Method 2). Simulations for selected catchments were carried out with a daily time step for the period 1981-2010 using a 500 × 500 m spatial resolution. Catchment-scale NIR ranged between 0 and 4.3 million m3 and 0 and 7.3 million m3 for the two methods, respectively, with no trend over the observation period in any catchment. During the heat wave in 2003, NIR increased by a factor of 1.5 to 2.3 relative to the mean, and in catchments where discharge is directly dependent on precipitation, NIR in the summer of 2003 reached the limits of river water availability. In contrast, in a region with water supply from glacier melt water, highest NIR in 2003 still remained far below total river discharge. The results show that NIR varies strongly between years and across the landscape, and even in a presently cool-temperate climate, irrigation may put pressure on regional water resources under extreme climatic conditions that may become more frequent by the end of the 21st century.展开更多
基金supported by the Erasmus Mundus Action 2 Programme of the European Union and the German Weather Service(DWD)and the Czech Hydrological-Meteorological Service(CHMI).
文摘In this study, recurrent networks to downscale meteorological fields of the ERA-40 re-analysis dataset with focus on the meso-scale water balance were investigated. Therefore two types of recurrent neural networks were used. The first approach is a coupling between a recurrent neural network and a distributed watershed model and the second a nonlinear autoregressive with exogenous inputs (NARX) network, which directly predicted the component of the water balance. The approaches were deployed for a meso-scale catchment area in the Free State of Saxony, Germany. The results show that the coupled approach did not perform as well as the NARX network. But the meteorological output of the coupled approach already reaches an adequate quality. However the coupled model generates as input for the watershed model insufficient daily precipitation sums and not enough wet days were predicted. Hence the long-term annual cycle of the water balance could not be preserved with acceptable quality in contrary to the NARX approach. The residual storage change term indicates physical restrictions of the plausibility of the neural networks, whereas the physically based correlations among?the components of the water balance were preserved more accurately by the coupled approach.
文摘With climate change, water may become limited for intensive agriculture even in regions presently considered “water-rich”. Information about the potential water requirement and its temporal and spatial variability can help to develop future water management plans. A case study was carried out for Switzerland with its highly complex pre-alpine topography and steep gradients in climate. The hydrological model WaSiM-ETH was used to simulate net irrigation requirement (NIR) for cropland, grassland and orchards using criteria to define irrigation periods based either on the water stress level (expressed by the ratio of actual (aET) to potential evapotranspiration ((pET) (Method 1) or on thresholds for soil water potential (Method 2). Simulations for selected catchments were carried out with a daily time step for the period 1981-2010 using a 500 × 500 m spatial resolution. Catchment-scale NIR ranged between 0 and 4.3 million m3 and 0 and 7.3 million m3 for the two methods, respectively, with no trend over the observation period in any catchment. During the heat wave in 2003, NIR increased by a factor of 1.5 to 2.3 relative to the mean, and in catchments where discharge is directly dependent on precipitation, NIR in the summer of 2003 reached the limits of river water availability. In contrast, in a region with water supply from glacier melt water, highest NIR in 2003 still remained far below total river discharge. The results show that NIR varies strongly between years and across the landscape, and even in a presently cool-temperate climate, irrigation may put pressure on regional water resources under extreme climatic conditions that may become more frequent by the end of the 21st century.
