基于BP神经网络的鄱阳湖水位模拟

考虑到鄱阳湖水位受流域五河与长江来水等多因素的共同作用而表现出高度非线性响应,采用典型的三层BPNN神经网络模型来模拟鄱阳湖水位与其主控因子之间的响应关系。分别将湖口、星子、都昌、棠荫和康山水位作为目标变量进行BPNN模型构建和适用性评估。结果显示：综合考虑流域五河及长江来水（汉口或九江）的BPNN水位模型,空间站点水位模拟精度（R2和Ens）可达090以上,各站点的均方根误差（RMSE）变化范围约050～10 m,若忽略长江来水的影响作用,仅将流域五河来水作为湖泊水位的主控影响因子,模型训练期与测试期的纳希效率系数（Ens）和确定性系数（R2）显著降低,且低于050,均方根误差（RMSE）也明显增大（124～288 m）,意味着综合考虑流域五河与长江来水是获取结构合理、精度保证的鄱阳湖水位模型的重要前提。同时建议针对鄱阳湖湖盆变化对水位的影响,尽可能选择一致性较好的长序列数据集来训练和测试BPNN模型。所构建的BPNN神经网络模型可进一步结合流域水文模型,用来预测气候变化与人类活动下流域径流变化对湖泊水位的潜在影响,也可作为一种有效的模型工具来回答当前鄱阳湖一些备受关注的热点问题,如定量区分流域五河与长江来水对湖泊洪枯水位的贡献分量,为湖泊洪涝灾害的防治和对策制定提供科学依据     

INTEGRATED ECHO SOUNDER,GPS, AND GIS FOR RESERVOIR SEDIMENTATION STUDIES: EXAMPLES FROM TWO ARKANSAS LAKES1

ABSTRACT: Bathymetric and sedimentation surveys were conducted using a dual frequency (28/200 kHz) echo sounder system in two reservoirs (Lee Creek Reservoir and Lake Shepherd Springs) in the Ozark Plateau of northwestern Arkansas. Echo sounder survey data were merged within geographic information system (GIS) software to provide detailed visualization and analyses of current depths, pre‐impoundment topography, distribution, thickness, and volume estimates of lacustrine sediment, time averaged sediment accumulation rates, long term average annual sediment flux, and water storage capacity. Calculated long term average sediment accumulation rates were used to model sediment infilling and projected lifetimes of each reservoir. Results from echo sounder surveys and GIS analyses suggest that the Lee Creek Reservoir has a projected lifetime of approximately 500 years compared to a projected lifetime for Lake Shepherd Springs of approximately 3,000 years. Estimated differences in projected lifetimes of these reservoirs reflected differences in initial reservoir volume and long term average annual sediment flux from the respective watersheds related to watershed area, physiography, land cover, and land use. The universal soil loss equation (USLE) model generated sediment fluxes an order of magnitude larger from the watersheds of both reservoirs compared to the geophysical data estimates. This study demonstrated the utility of merging geophysical survey (echo sounder) data within a GIS as an aid to understanding patterns of reservoir sedimentation. These data and analyses also provide a baseline relevant to understanding sedimentation processes and are necessary for development of long term management plans for these reservoirs and their watersheds.     

OPTIMAL BATHYMETRY OF TYRRHENIAN SEA BY STOCHASTIC INTERPOLATORS1

ABSTRACT: The analysis of the sea depth in the vicinity of a thermal power plant is important in view of the heat dispersion and transport of the cooling water discharged from the energy production processes. The bathymetry of the Tyrrhenian Sea in the area facing the thermoelectric plant being built at Montalto di Castro, Italy, is reconstructed with the aid of the kriging technique. The large number of measurements (2038) precludes the use of a global interpolation model. At the same time, the steepness of the variogram suggests that the stochastic process is likely non-stationary. The latter problem is managed by detrending the observed records with a fourth-order polynomial drift determined with the least squares method. The global approach is replaced with local models with an influence radius of 70 m and a minimal number of interpolation records equal to 10. The detrended residuals appear to be second-order stationary having an isotropic vaniogram that is fitted equally well by a linear or a spherical function. The average size of the local interpolation models is 28, which ensures a high cemputational efficiency. The bathymetry of the sea area of interest is then reconstructed over a regular grid with a total of 3570 nodes. The estimation standard deviation maps show that the accuracy of the prediction is even greater than the prescribed tolerance (= 0.2 m) and indicate that future data collection campaigns require a less dense and, therefore, less expensive recording network. The redundant measurement profiles are identified. Their elimination still allows for bathymetry estimates affected by interpolation errors that fall within the project specifications.     

Bottomfast Ice Mapping and the Measurement of Ice Thickness on Tundra Lakes Using C‐Band Synthetic Aperture Radar Remote Sensing1

Abstract: Industrial activity in Canada’s north is increasing, placing demands on the use of water from lakes to build ice roads. Winter water withdrawal from these lakes has the potential to impact overwintering fish. Removal of water from small lakes can decrease oxygen and habitat available to fish. To address this issue, a protocol has been developed by the Department of Fisheries and Oceans outlining water withdrawal thresholds. Bathymetric surveys are the traditional method to determine lake depth, but are costly given the remoteness of northern lakes. This paper investigates the use of satellite C‐band synthetic aperture radar (SAR) remote sensing technology as a potential alternative or complement to traditional survey methods. Previous research has shown that a SAR can detect the transition from grounded to floating ice on lakes, or if a lake is completely frozen. Grounded ice has a dark signature while floating ice appears very bright in contrast. Similar results were observed for the datasets acquired in the study area. This suggests that lakes that freeze completely to the bottom can be identified using SAR. Such water bodies would not be viable fish overwintering habitat and can therefore be used as water sources without thresholds necessary. However, attempts to accurately calculate the depth of the ice at the grounded‐floating ice boundary using bathymetric profiles acquired in the summer and lake ice thickness measurements from a reference lake near Inuvik proved to be unreliable.     

Representation of the world's biophysical conditions by the global protected area network

Protected areas (PAs) are often implemented without consideration of already existing PAs, which is likely to cause an overrepresentation of certain biophysical conditions. We assessed the representativeness of the current PA network with regard to the world's biophysical conditions to highlight which conditions are underprotected and where these conditions are located. We overlaid terrestrial and marine PAs with information on biophysical conditions (e.g., temperature, precipitation, and elevation) and then quantified the percentage of area covered by the PA network. For 1 variable at a time in the terrestrial realm, high temperature, low precipitation, and medium and very high elevation were underrepresented. For the marine realm, low and medium sea surface temperature (SST), medium and high sea surface salinity (SSS), and the deep sea were underrepresented. Overall, protection was evenly distributed for elevation across the terrestrial realm and SST across the marine realm. For 2 variables at a time, cold and very dry terrestrial environments had mostly low protection, which was also the case for low SST and low and medium SSS across most depths for marine environments. Low protection occurred mostly in the Sahara and the Arabian Peninsula for the terrestrial realm and along the Tropic of Capricorn and toward the poles for the marine realm. Although biodiversity measures are of prime importance for the design of PA networks, highlighting biophysical gaps in current PAs adds a frequently overlooked perspective. These gaps may weaken the potential of PAs to conserve biodiversity. Thus, our results may provide useful insights for researchers, practitioners, and policy makers to establish a more comprehensive global PA network.