Advancing real-time error correction of flood forecasting based on the hydrologic similarity theory and machine learning techniques.

Real-time flood forecasting is one of the most pivotal measures for flood management, and real-time error correction is a critical step to guarantee the reliability of forecasting results. However, it is still challenging to develop a robust error correction technique due to the limited cognitions of catchment mechanisms and multi-source errors across hydrological modeling. In this study, we proposed a hydrologic similarity-based correction (HSBC) framework, which hybridizes hydrological modeling and multiple machine learning algorithms to advance the error correction of real-time flood forecasting. This framework can quickly and accurately retrieve similar historical simulation errors for different types of real-time floods by integrating clustering, supervised classification, and similarity retrieval methods. The simulation errors "carried" by similar historical floods are extracted to update the real-time forecasting results. Here, combining the Xin'anjiang model-based forecasting platform with k-means, K-nearest neighbor (KNN), and embedding based subsequences matching (EBSM) method, we constructed the HSBC framework and applied it to China's Dufengkeng Basin. Three schemes, including "non-corrected" (scheme 1), "auto-regressive (AR) corrected" (scheme 2), and "HSBC corrected" (scheme 3), were built for comparison purpose. The results indicated the following: 1) the proposed framework can successfully retrieval similar simulation errors with a considerable retrieval accuracy (2.79) and time consumption (228.18 s). 2) four evaluation metrics indicated that the HSBC-based scheme 3 performed much better than the AR-based scheme 2 in terms of both the whole flood process and the peak discharge; 3) the proposed framework overcame the shortcoming of the AR model that have poor correction for the flood peaks and can provide more significant correction for the floods with bad forecasting performance. Overall, the HSBC framework demonstrates the advancement of benefiting the real-time error correction from hydrologic similarity theory and provides a novel methodological alternative for flood control and water management in wider areas.

A continuation-dynamic constitution analysis approach based on digital stable marker tracing and study on simulation of ecological tidal water diversion.

Water Diversion Projects have become increasingly popular in improving water quality in various water ecosystems. However, these projects also require a more comprehensive evaluation. In this study, we introduced a digital stable marker tracing module and proposed a continuation-dynamic constitution analysis approach. We applied this approach to analyze the ecological tidal water diversion in Changshu town, China. The results showed that the mean diversion water age of the Yangtze River water source was 10.80 h, the residence time of the background water source in Baimaotang was approximately 4.0 h, and the contribution of inflow water sources from tributaries accounted for 15% of discharges. The results can demonstrate practicality of our approach in quantitatively evaluating water diversion impacts and optimizing cooperative diversion projects. Furthermore, our discussion led to the design of an ecological tidal water diversion based on optimized cooperative diversion, which showed element-complementary and whole-comprehensive effects. This indicates that the ecological tidal water diversion can extend the impact of cooperative diversion. The continuation-dynamic constitution analysis approach enhances the tracing capacity of inflow constitution and enables the distinction of different time-varying distributions of each inflow constitution. Therefore, this approach holds promise as an embedded "Digital stable marker tracing" module in the model.

Establishment of watershed ecological water requirements framework: A case study of the Lower Yellow River, China.

It is of great practical significance to ensure ecological water requirements (EWRs) for the maintenance of river health and the sustainable development of human socioeconomics. How to scientifically determine the comprehensive EWRs and estimate the uncertainty of hydro-ecological tools performed in the process of conducting remains one of the most important yet most complicated issues. In this study, the ecological water requirements framework (EWRsF) of the Lower Yellow River (LYR), which considers instream ecological base flow, survival and reproduction of indicator fish species, equilibrium of erosion and siltation and ecological function of the estuary, was constructed by integrating hydrological, hydraulic and ecological habitat methods. The framework contains three crucial components - determination of instream EWRs and estuarine EWRs, uncertainty analysis of hydro-ecological tools. For instream ecological base flow, we proposed an improved Tennant method, which took into account both seasonality and sediment transport characteristics of the LYR, and could better reflect the actual hydrological regime. For the hydrological ecological response relationship of indicator fish species, we estimated the uncertainty of the model output of River2D to improve its credibility of the simulation results. The results demonstrated that: 1) Two-grade intra-annual monthly EWRs process of suitable and minimum for four instream sections and estuary area were obtained. The flood season (June-October) is the period with the largest proportion of intra-annual instream EWRs, whereas in estuary area, is the spawning period (April-July) of dominant species. 2) The uncertainty of HSI curves directly leads to the uncertainty of model output. Although the shape and position of the WUA curve can be uncertain, it does not affect the judgment of EWRs threshold. 3) The research results can provide scientific basis for water resource management decision-making in the LYR. Additionally, the ideas also have reference significance for similar basins.

Nonstationary flood coincidence risk analysis using time-varying copula functions.

The coincidence of flood flows in a mainstream and its tributaries may lead to catastrophic floods. In this paper, we investigated the flood coincidence risk under nonstationary conditions arising from climate changes. The coincidence probabilities considering flood occurrence dates and flood magnitudes were calculated using nonstationary multivariate models and compared with those from stationary models. In addition, the "most likely" design based on copula theory was used to provide the most likely flood coincidence scenarios. The Huai River and Hong River were selected as case studies. The results show that the highest probabilities of flood coincidence occur in mid-July. The marginal distributions for the flood magnitudes of the two rivers are nonstationary, and time-varying copulas provide a better fit than stationary copulas for the dependence structure of the flood magnitudes. Considering the annual coincidence probabilities for given flood magnitudes and the "most likely" design, the stationary model may underestimate the risk of flood coincidence in wet years or overestimate this risk in dry years. Therefore, it is necessary to use nonstationary models in climate change scenarios.

New method to calculate the dynamic factor-flow velocity in Geomorphologic instantaneous unit hydrograph.

The determination of characteristic flow velocity is a hydrodynamic problem needs to be solved in the application of geomorphologic instantaneous unit hydrograph (GIUH) for runoff simulation in areas with no or limited data. In this study, 120 watersheds are collected to construct a regression model; 85 of these basins are used for regression analysis, and the 35 remaining basins are utilized to verify the feasibility of the constructed model. Random forest algorithm is applied to screen out important geomorphologic factors from the 16 extracted factors that may affect flow velocity. Multivariate regression is used to establish the numerical relationship between velocity and the selected factors. Sensitivity analysis of each adopted factor in the constructed model is conducted using the LH-OAT method. The rationality and feasibility of the regression model are validated by comparing the flow velocity calculation with a previous approach, which is also calculated based on geomorphological parameters. Subsequently, the runoff simulation based on the GIUH model is evaluated using the proposed technique. Results demonstrate that the proposed formula possesses high fitting accuracy and can be easily used to calculate flow velocity and generate GIUH.

Temporal 18O and deuterium variations in hydrologic components of a small watershed during a typhoon event.

The differences between δ18O and δ2H in throughfall and open rainfall were studied for a selected typhoon event in a watershed within the Taihu Lake drainage basin, eastern China. In this event, the isotopic composition of precipitation exhibited a strong temporal variation. Comparison results show that an isotopic composition difference existed not only between gross rainfall and average incremental rainfall, but also between different calculation methods used to derive average. The differences between incremental precipitation and throughfall isotopic composition were observed in this study. Considering the temporal variation in rainfall and throughfall during this typhoon event, the incremental value can have an effect on hydrograph separation more accurately in evaluating the importance of 'new' water. In addition, isotopic fluctuations of surface water and groundwater differed from those of rainfall and throughfall throughout the event.

Differences in oxygen-18 and deuterium content of throughfall and rainfall during different flood events in a small headwater watershed.

Inter-storm stable isotopic values of rainfall and throughfall for three flooding events were measured during the period of July to August 2011 to estimate their differences in a first-order chestnut watershed, Meilin, within the Taihu Lake basin. Comparison of δ(2)H and δ(18)O was conducted from four aspects: (1) sampling methods, (2) calculation methods, (3) stable isotopes and (4) flood events. Arithmetic and weighted incremental values of throughfall were generally lighter than those of rainfall. Isotopic composition of both incremental rainfall and throughfall exhibits marked temporal variation, particularly during large storm events; the former shows a higher variation than the latter. Differences of averaged precipitation and throughfall between storms were small, but individual storm variations were larger. Isotopic differences using different calculation methods are not significant; however, the differences resulting from sampling methods are of greater importance.