The evaluation of surface water resources is a necessary input to solving water management problems. Neural network models have been trained to predict monthly runoff for the Tirso basin, located in Sardinia (Italy) at the S. Chiara section. Monthly time series data were available for 69 years and are characterized by non-stationarity and seasonal irregularity, which is typical of a Mediterranean weather regime. This paper investigates the effects of data preprocessing on model performance using continuous and discrete wavelet transforms and data partitioning. The results showed that networks trained with pre-processed data performed better than networks trained on undecomposed, noisy raw signals. In particular, the best results were obtained using the data partitioning technique. 相似文献
We investigated the effectiveness of combining gauge observations and satellite-derived precipitation on flood forecasting. Two data merging processes were proposed: the first one assumes that the individual precipitation measurement is non-bias, while the second process assumes that each precipitation source is biased and both weighting factor and bias parameters are to be calculated. Best weighting factors as well as the bias parameters were calculated by minimizing the error of hourly runoff prediction over Wu-Tu watershed in Taiwan. To simulate the hydrologic response from various sources of rainfall sequences, in our experiment, a recurrent neural network (RNN) model was used.
The results demonstrate that the merged method used in this study can efficiently combine the information from both rainfall sources to improve the accuracy of flood forecasting during typhoon periods. The contribution of satellite-based rainfall, being represented by the weighting factor, to the merging product, however, is highly related to the effectiveness of ground-based rainfall observation provided gauged. As the number of gauge observations in the basin is increased, the effectiveness of satellite-based observation to the merged rainfall is reduced. This is because the gauge measurements provide sufficient information for flood forecasting; as a result the improvements added on satellite-based rainfall are limited. This study provides a potential advantage for extending satellite-derived precipitation to those watersheds where gauge observations are limited. 相似文献
Statistical methods are widely applied in visibility forecasting. In this article, further improvements are explored by extending
the standard probabilistic neural network approach. The first approach is to use several models to obtain an averaged output,
instead of just selecting the overall best one, while the second approach is to use deterministic neural networks to make
input variables for the probabilistic neural network. These approaches are extensively tested at two sites and seen to improve
upon the standard approach, although the improvements for one of the sites were not found to be of statistical significance. 相似文献
The scientific community that includes meteorologists, physical scientists, engineers, medical doctors, biologists, and environmentalists
has shown interest in a better understanding of fog for years because of its effects on, directly or indirectly, the daily
life of human beings. The total economic losses associated with the impact of the presence of fog on aviation, marine and
land transportation can be comparable to those of tornadoes or, in some cases, winter storms and hurricanes. The number of
articles including the word ``fog' in Journals of American Meteorological Society alone was found to be about 4700, indicating
that there is substantial interest in this subject. In spite of this extensive body of work, our ability to accurately forecast/nowcast
fog remains limited due to our incomplete understanding of the fog processes over various time and space scales. Fog processes
involve droplet microphysics, aerosol chemistry, radiation, turbulence, large/small-scale dynamics, and surface conditions
(e.g., partaining to the presence of ice, snow, liquid, plants, and various types of soil). This review paper summarizes past
achievements related to the understanding of fog formation, development and decay, and in this respect, the analysis of observations
and the development of forecasting models and remote sensing methods are discussed in detail. Finally, future perspectives
for fog-related research are highlighted. 相似文献
The role of the North Atlantic Oscillation (NAO) in effecting changes in winter extreme high and low waters and storm surges in UK waters has been investigated with the use of a depth-averaged tide+surge numerical model. Spatial patterns of correlation of extreme high and low waters (extreme still water sea levels) with the NAO index are similar to those of median or mean sea level studied previously. Explanations for the similarities, and for differences where they occur, are proposed. Spatial patterns of correlations of extreme high and low and median surge with the NAO index are similar to the corresponding extreme sea-level patterns. Suggestions are made as to which properties of surges (frequency, duration, magnitude) are linked most closely to NAO variability. Several climate models suggest higher (more positive) average values of NAO index during the next 100 years. However, the impact on the UK coastline in terms of increased flood risk should be low (aside from other consequences of climate change such as a global sea-level rise) if the existing relationships between extreme high waters and NAO index are maintained. 相似文献