Actual Rainfall Events Research Articles

Enhancing flood forecasting and warning precision through multi-task deep learning approaches

ABSTRACT This study proposes a multi-task deep learning model for simultaneous prediction of time-series water levels and flood risk thresholds, aiming to enhance flood forecasting precision. Using AutoKeras, single-task and multi-task models were optimised to predict water levels 10–360 min ahead based on 720 min of prior data. The multi-task model consistently outperformed the single-task model across multiple evaluation metrics, including correlation coefficients, RMSE, Nash–Sutcliffe efficiency, and Kling–Gupta efficiency scores. Real-time prediction tests on actual rainfall events further validated the multi-task model's improved accuracy and applicability in operational flood forecasting. The study demonstrates significant progress in flood prediction methodologies, offering a more comprehensive approach to forecasting and categorising flood incidents.

Clarifying urban flood response characteristics and improving interpretable flood prediction with sparse data considering the coupling effect of rainfall and drainage pipeline siltation

With climate warming and accelerated urbanisation, severe urban flooding has become a common problem worldwide. Frequent extreme rainfall events and the siltation of drainage pipes further increase the burden on urban drainage networks. However, existing studies have not fully considered the effects of rainfall and pipeline siltation on the response characteristics of flooding when constructing numerical models of urban flooding simulations. To solve this problem, a surface-subsurface coupling model was constructed by combining the Saint-Venant equation, Manning equation, a one-dimensional pipeline model (SWMM), and a two-dimensional surface overflow model (LISFLOOD-FP). Then, the SWMM model considering pipeline siltation and the two-dimensional surface overflow model (LISFLOOD-FP) were coupled with the flow exchange governing equation, and the urban flooding response characteristics considering the coupling effect of “rainfall and drainage pipeline siltation” were analysed. To enhance the solvability of waterlogging prediction, an intelligent prediction model of urban flooding based on Bayes-CNN-BLSTM was established by combining a convolutional neural network (CNN), bidirectional long short-term memory neural network (BLSTM), Bayesian optimisation (Bayes), and an interpretable loss function error correction method. The actual rainfall events and flooding processes recorded by the monitoring equipment at Huizhou University were used to calibrate and verify the model. The results show that in the Rainfall 1 and Rainfall 2 scenarios, the overload rates of the pipelines in the current siltation scenario were 60.06 % and 68.37 %, respectively, and the proportions of overflow nodes were 24.87 % and 25.89 %, respectively. When the drainage network was initially put into operation, the overload rates of the pipeline were 36.67 % and 41.16 %, and the overflow nodes accounted for 3.05 % and 4.06 %, respectively. The inundated area and volume of urban flooding increased when the combined siltation coefficient (CSC) was 0.2; therefore, two desilting schemes were determined. Under Rainfall 1, Rainfall 2, and the four rainfall recurrence periods, the Bayes-CNN-BLSTM model had clear advantages in terms of accuracy, reliability, and robustness.

Evaluation of GPM IMERG Satellite Precipitation Products in Event-Based Flood Modeling over the Sunshui River Basin in Southwestern China

This study evaluates the applicability of hourly Global Precipitation Measurement Mission (GPM) Integrated Multi-satellitE Retrievals for GPM (IMERG) data for event-based flood modeling in the Sunshui River Basin, southwestern China, using the hydrologic modeling system (HEC-HMS) model. The accuracies of IMERG V6, IMERG V7, and the corrected IMERG V7 satellite precipitation products (SPPs) were assessed against ground rainfall observations. The performance of flood modeling based on the original and the corrected SPPs was then evaluated and compared. In addition, the ability of different numbers (one–eight) of ground stations to correct IMERG V7 data for flood modeling was investigated. The results indicate that IMERG V6 data generally underestimate the actual rainfall of the study area, while IMERG V7 and the corrected IMERG V7 data using the geographical discrepancy analysis (GDA) method overestimate rainfall. The corrected IMERG V7 data performed best in capturing the actual rainfall events, followed by IMERG V7 and IMERG V6 data, respectively. The IMERG V7-generated flood hydrographs exhibited the same trend as those of the measured data, yet the former generally overestimated the flood peak due to its overestimation of rainfall. The corrected IMERG V7 data led to superior event-based flood modeling performance compared to the other datasets. Furthermore, when the number of ground stations used to correct the IMERG V7 data in the study area was greater than or equal to four, the flood modeling performance was satisfactory. The results confirm the applicability of IMERG V7 data for fine time scales in event-based flood modeling and reveal that using the GDA method to correct SPPs can greatly enhance the accuracy of flood modeling. This study can act as a basis for flood research in data-scarce areas.

Assessing the accuracy of large-scale rainfall erosivity estimation based on climate zones and rainfall patterns

Rainfall erosivity is affected by the amount and intensity of rainfall in a certain period, which is an essential factor for soil erosion prediction. However, it is generally calculated by field measurements on a local scale. With a focus on global soil erosion assessment, some researchers have improved the estimation of global rainfall erosivity by using statistical models in some climate zones. However, the climate zones cannot represent actual erosive rainfall events. Therefore, such usage of models would lead to more uncertainties when estimating rainfall erosivity across the globe. Here, our study compared six common-used models of rainfall erosivity and then improved the accuracy of rainfall erosivity estimations based on global rainfall patterns, which are defined by the amount and distribution of rainfall in a year. Results showed that: (1) Compared with the climate zone classification, the model fitting under the rainfall pattern classification can improve the model accuracy and result in higher variation among the rainfall patterns. The average accuracy of all models was improved by 8%, and the accuracy of annual models was increased by 33%. (2) Models based on annual rainfall are more suitable for the drought and seasonal rainfall patterns, while most models based on monthly rainfall are suitable for the moderate rainfall pattern. However, most models based on monthly or annual rainfall have high uncertainties and low accuracies in regions with annual precipitation < 200 mm or > 850 mm. This study can provide helpful implications for model selection and parameter calibration associated with large-scale water erosion prediction.

Combining rainfall-induced shallow landslides and subsequent debris flows for hazard chain prediction

Landslides, debris flows, and other destructive natural hazards induced by heavy rainfall in mountainous regions are sometimes not independent but combined to form a disaster chain. Based on the integral link between the triggering of the landslide and the subsequent debris flow, we propose an approach that combines the Transient Rainfall Infiltration and Grid-Based Regional Slope Stability (TRIGRS) model and the Rapid Mass Movements Simulation (RAMMS) model to achieve hourly hazard prediction. The results indicate that the TRIGRS model performed well in predicting the spatial distribution of the shallow landslides, with a success rate of 81.86%. Thus, it is reasonable to use it as the initial input for debris flow simulations. The relationship between the landslide area and the accumulated rainfall obtained using the TRIGRS model is a power-law relationship, which provides a reference for regions that lack rainfall data to predict the material source of a debris flow. The coupled model was found to have a good accuracy of 76.77% in simulating the debris flow. This was close to the debris flow simulation based on the interpreted landslides, and it still produced reasonable results and a more practical value. Furthermore, the proposed coupled model can dynamically predict disasters by the hour based on actual rainfall events. Therefore, the results of this study help provide a more complete hazard prediction picture for rainfall-induced landslide-debris flow hazards in mountainous regions.

Modeling the quality of sewage during the leaking of stormwater surface runoff to the sanitary sewer system using SWMM: a case study

Abstract This paper describes the application of the storm water management model (SWMM) for predicting the sewage quality in the sanitary sewer system of the study area resulting from the leaking of stormwater surface runoff to the system during rainfall events at different return periods. The concentrations of major pollutants were assessed in the sanitary sewer system at different rainfall intensities. Then, a solution to mitigate the problem was proposed using low impact development (LID) technology. The results of sensitivity analysis indicated that maximum build-up possible was the most sensitive parameter for model calibration. The model was calibrated using actual rainfall events, and statistical validation coefficients of R (0.81–0.82) and NMSE (0.0173–0.022) proved that the model is valid. The sewage quality assessment results showed that pollutants concentration increased to its maximum level at 20 min and gradually decreased to a slightly constant minimum value after 2 h. The proposed solution of LID reduced the pollutants concentrations by 82–88, 75–77, 52–55, and 7–10% for all pollutants at return periods of 2, 5, 10, and 25 years, respectively. To conclude, SWMM simulation successfully predicted the concentration of the pollutants, and leaking of stormwater surface runoff has changed the sewage quality.

Evaluation of hydrological modeling using climatic station and gridded precipitation dataset

The conventional rainfall data estimates are relatively accurate at some points of the region. The interpolation of such type of data approximates the actual rainfield however in data scarce regions; the resulted rainfield is the rough estimate of the actual rainfall events. In data scarce regions like Indus basin Pakistan, the data obtained through remote sensing can be very useful. This research evaluates two types of gridded data i.e., European Reanalysis (ERA) interim and Japanese Reanalysis 55 years (JRA-55) along with the climatic station data for three small dams in Pakistan. Since no measured flow data is available at these dams, the nearest possible catchments where flow data is available are calibrated and the calibrated parameters of these catchments are then used in actual dams for simulating the flow from all the three types of data using Soil and Water Assessment Tool (SWAT). The results of the comparison of gridded and rainguage precipitation shows that gridded data highly overestimates the climatic station data. Similar results were observed in the comparison of flow simulated by SWAT model. The Peak flood calculated from JRA-55 overestimates while the Era-Interim peak floods are comparable to that of climatic stations in two of the three catchments.

Using the General Regression Neural Network Method to Calibrate the Parameters of a Sub-Catchment

Computer software is an effective tool for simulating urban rainfall–runoff. In hydrological analyses, the storm water management model (SWMM) is widely used throughout the world. However, this model is ineffective for parameter calibration and verification owing to the complexity associated with monitoring data onsite. In the present study, the general regression neural network (GRNN) is used to predict the parameters of the catchment directly, which cannot be achieved using SWMM. Then, the runoff curve is simulated using SWMM, employing predicted parameters based on actual rainfall events. Finally, the simulated and observed runoff curves are compared. The results demonstrate that using GRNN to predict parameters is helpful for achieving simulation results with high accuracy. Thus, combining GRNN and SWMM creates an effective tool for rainfall–runoff simulation.

Calibration of High-Impact Short-Range Quantitative Precipitation Forecast through Frequency-Matching Techniques

Short-lived and intense rainfall is common in Hong Kong during the wet season and often causes disruption to daily lives. As global numerical weather prediction (NWP) models are progressively improved, representation of sub-synoptic or mesoscale systems associated with intense rainfall is better parametrized and resolved than ever before, but their quantitative precipitation forecasts (QPFs) still tend to underestimate the magnitude of intense rainfall. This study calibrated model QPFs over the region of Hong Kong by two frequency-matching methods. In both methods, conversion schemes between the direct model output (DMO) and calibrated forecasts were first established by matching the cumulative distribution to that of the observed data. The “Adaptive Table” method updated the conversion scheme whenever the latest observation fell out of its expected range in the existing scheme, whereas the “Sliding Window” method reconstructed the conversion scheme using data from the most recent two years. The calibration methods had been verified against actual rainfall events with different thresholds, and it was found that both methods could improve model performance for moderate and heavy rainfall in short-range forecasts with similar effectiveness. They were also able to reduce the systematic bias of precipitation forecasts for significant rainfall throughout the verification period.

Experimental Study on the Retention and Interception Effect of an Extensive Green Roof (GR) with a Substrate Layer Modified with Kaolin

Extensive green roofs (GRs) often appear as pollution sources during actual rainfall events; therefore, it is necessary to study the control of nutrient leaching in the substrate layer. In this study, four extensive GR experimental devices are built: two with artificial granular structure substrate layers improved with kaolin as a binder, one with a commercial substrate layer, and one with a standard roof (SR). Based on the simulated rainfall conditions in different local recurrence periods, the delayed outflow time, rainfall retention rate, event mean concentration (EMC), and cumulative pollutant quality of NH4+, NO3−, NO2−, and PO43− in the effluents were measured and evaluated. The results of the study indicate that under simulated rainfall in all the experimental design recurrence periods, the kaolin-modified substrate layer does not exhibit a more significant retention capacity than the commercial substrate. However, it does show some suppression of the leaching effect of NO3− and PO43− in the runoff. The reduction rate of cumulative NO3− quality is 6.56%, and PO43− is 10.54%. In future practical engineering and related research, attention should be paid to the influence of the type and addition amount of the substrate layer modifier on the stability of the granular structure to prevent nutrient loss caused by soil erosion.

The performance of the IMERG satellite-based product in identifying sub-daily rainfall events and their properties

Sub-daily rainfall information is essential for many hydrological applications, but ground-based data availability is still an issue in poorly gauged regions worldwide. Satellite remote sensing missions, such as the Global Precipitation Measurement (GPM) mission, have been playing a key role in estimating sub-daily rainfall data globally. However, the quality of such information needs to be carefully evaluated. Previous studies evaluating sub-daily data from the Integrated multi-satellitE Retrievals for GPM (IMERG) product considered only the rainfall depth over pre-defined periods (e.g., hourly or half-hourly), with no analysis of the ability and quality of the product in defining the actual rainfall events and the associated properties. Thus, the objective of this study is to evaluate the performance of the IMERG Final Run Version 06B (V06B) product in capturing sub-daily rainfall events and their properties (depth, duration and intensity) over Brazil. The analysis consisted of comparing the satellite estimates against the ground-based data from 1757 sub-daily rainfall gauges for a period of three years (2015–2017). This study used the minimum inter-event time (MIT) criterion to define independent rainfall events determined by dry periods: 1, 6 and 24 h. Results show that IMERG can properly estimate the sub-daily rainfall depth for the three MITs considered, with the best results found in the southern part of the country. This means that the IMERG product represents a good source of sub-daily rainfall depth data for hydrological and hydroclimatic applications in Brazil. On the other hand, the evaluation shows large overestimations and underestimations of the IMERG product for rainfall duration and intensity properties, respectively. The results obtained from this study provide a reference for IMERG users who require sub-daily rainfall data based on events and further knowledge about its properties.

An Improved Simplified Urban Storm Inundation Model Based on Urban Terrain and Catchment Modification

Flooding caused by unpredictable high-intensity rainfall events in urban areas has become a global phenomenon due to the combined effect of urbanization and climate change. There are numerous hydrodynamic models for urban flooding simulation and management. However, it is difficult for most of these models to simplify the surface runoff process and still provide high simulation accuracy. In this study, an improved simplified urban storm inundation model (SUSIM) that integrates urban terrain, precipitation, surface runoff and inundation models was proposed to quickly and accurately simulate the different inundation conditions by modifying the urban terrain and catchments. Haining City, China, was selected as a case study in which SUSIM was tested and validated. The results were as follows: (1) Detailed locations and depths of inundation were quickly calculated with high correlation coefficient (≥75%) compared to three actual rainfall events. (2) Four scenarios under different rainfall intensities (5-, 10-, 20- and 50-year return period, respectively) were designed. The maximum inundation depths significantly increased from 403 mm to 1522 mm and the maximum inundation area increased from 2904 m2 to 7330 m2. According to the simulation results, Haining Avenue, the West Mountain Park and the old urban area in the northeast part of the city would encounter the most extensive and severe inundation. The result reveals that the SUSIM could find inundation locations and calculate inundation depth and area quickly. It provides better insights and tools for urban inundation simulation and planning strategies.

Probability distribution analysis of extreme rainfall events in a flood-prone region of Mumbai, India

Mumbai, being a coastal city surrounded by sea and creeks, is vulnerable to severe flooding almost every year owing to intense rainfall. Therefore, it is vital to quantify the magnitude of extreme rainfall events, which mainly depends upon the suitability of the derived distribution and the selection of a technique for processing and analysing data. The magnitude is often estimated using window-based approach of rainfall, which does not represent actual rainfall events since the duration of the rainfall is prefixed in this approach. Hence, the goal is to determine the best-fit probability distribution for extreme storms using the ‘storm event analysis’ method to overcome this limitation. The distribution is analysed with two-parameter [Lognormal (LN), Gamma (G), Gumbel (GUM) and Weibull (W)] and three-parameter [generalised extreme value (GEV), generalised Pareto (GP), Log Pearson 3 (LP3) and Frechet (F)] distribution functions on two sets [maximum extreme volume (MEV) and the maximum peak intensity (MPI)] of extreme events. Goodness-of-fit tests are applied to the distribution functions of the extreme storm series (2006–2016) of 26 meteorological stations. The best score results show that the GEV distribution fits best for 29% of the stations, and 27% and 22% of the stations showed the best fit for the F and GP distributions, respectively. On a basin scale, MEV events are best described by the GEV distribution while GP or F distributions show the best fits for MPI events. However, each station must be analysed individually as none of GEV, GP or F is always the best-fitting distribution function.

Slope stability evaluation of an agricultural embankment by statistically derived rainfall patterns

Torrential rainfall has a detrimental impact on the stability of soil-based agricultural embankments due to seepage and slope failure. The impact on the stability of an embankment due to rainfall patterns is specific to the geography, topography, and geometry of the embankment. However, this stability has not been widely assessed and updated with respect to new climate regimes. In this study, the slope stability of an agricultural embankment was evaluated based on five synthetic rainfall patterns derived from actual heavy rainfall events. The rainfall patterns were generated based on common patterns of rainfall hyetographs used in hydrological design and practice. The stability of the embankment was computed using the limit equilibrium method. Results of the stability analyses revealed time-series slope instability under rainfall patterns, and particularly the rainfall pattern that caused the most instability to the embankment was identified. Therefore, reassessment of the stability of the embankment in potential areas prone to heavy rainfall and flooding is beneficial to the preparation of early warning and emergency planning and preparedness as countermeasures for mitigating sediment-related hazards and disasters.

Establishing a method of short-term rainfall forecasting based on GNSS-derived PWV and its application

AbstarctGlobal Navigation Satellite System (GNSS) can effectively retrieve precipitable water vapor (PWV) with high precision and high-temporal resolution. GNSS-derived PWV can be used to reflect water vapor variation in the process of strong convection weather. By studying the relationship between time-varying PWV and rainfall, it can be found that PWV contents increase sharply before raining. Therefore, a short-term rainfall forecasting method is proposed based on GNSS-derived PWV. Then the method is validated using hourly GNSS-PWV data from Zhejiang Continuously Operating Reference Station (CORS) network of the period 1 September 2014 to 31 August 2015 and its corresponding hourly rainfall information. The results show that the forecasted correct rate can reach about 80%, while the false alarm rate is about 66%. Compared with results of the previous studies, the correct rate is improved by about 7%, and the false alarm rate is comparable. The method is also applied to other three actual rainfall events of different regions, different durations, and different types. The results show that the method has good applicability and high accuracy, which can be used for rainfall forecasting, and in the future study, it can be assimilated with traditional weather forecasting techniques to improve the forecasted accuracy.

Independent Validation of the SWMM Green Roof Module

Green roofs are a popular Sustainable Drainage Systems (SuDS) technology. They provide multiple benefits, amongst which the retention of rainfall and detention of runoff are of particular interest to stormwater engineers. The hydrological performance of green roofs has been represented in various models, including the Storm Water Management Model (SWMM). The latest version of SWMM includes a new LID green roof module, which makes it possible to model the hydrological performance of a green roof by directly defining the physical parameters of a green roof’s three layers. However, to date, no study has validated the capability of this module for representing the hydrological performance of an extensive green roof in response to actual rainfall events. In this study, data from a previously-monitored extensive green roof test bed has been utilised to validate the SWMM green roof module for both long-term (173 events over a year) and short-term (per-event) simulations. With only 0.357% difference between measured and modelled annual retention, the uncalibrated model provided good estimates of total annual retention, but the modelled runoff depths deviated significantly from the measured data at certain times (particularly during summer) in the year. Retention results improved (with the difference between modelled and measured annual retention decreasing to 0.169% and the Nash-Sutcliffe Model Efficiency (NSME) coefficient for per-event rainfall depth reaching 0.948) when reductions in actual evapotranspiration due to reduced substrate moisture availability during prolonged dry conditions were used to provide revised estimates of monthly ET. However, this aspect of the model’s performance is ultimately limited by the failure to account for the influence of substrate moisture on actual ET rates. With significant differences existing between measured and simulated runoff and NSME coefficients of below 0.5, the uncalibrated model failed to provide reasonable predictions of the green roof’s detention performance, although this was significantly improved through calibration. To precisely model the hydrological behaviour of an extensive green roof with a plastic board drainage layer, some of the modelling structures in SWMM green roof module require further refinement.

Evaluation of error in TRMM 3B42V7 precipitation estimates over the Himalayan region

AbstractAccurate precipitation measurement is crucial for weather forecasting and hydrological modeling. Tropical Rainfall Measuring Mission (TRMM) 3B42V7 satellite precipitation product offers an opportunity to monitor precipitation at high spatiotemporal resolution. However, it has several inherent errors related to observation, instrument, and rainfall retrieval algorithms. It is, therefore, essential to validate it with ground‐based measurements. We divide the region into different elevation ranges and compare 3B42V7 with India Meteorological Department gauge‐based measurements, so as to observe the behavior of satellite at different altitudes. This paper evaluates error characteristics of 3B42V7 using continuous and categorical validation schemes. The analysis reveals 3100 m altitude as the breakpoint for the satellite overestimating and underestimating rainfall amount for elevation ranges below and above it, respectively. It gives a poor positive correlation of ~0.23 between individual rainfall events, though the correlation improves (~0.67) for areal‐averaged precipitation values. 3B42V7 also underestimates the frequency of actual rainfall events and is not very good at identifying correct rain and no‐rain events with the overall accuracy of ~66%. Conclusively, the satellite exhibits comparatively better performance for 1000–2000 m elevations but exacerbates over higher‐altitude regions. Further, we assess its capability for very heavy rainfall events using three percentile thresholds. The low‐magnitude bias for 98th and 99th percentiles and high‐magnitude bias for 99.99th percentile imply that 3B42V7 may not be suitable for the study of very heavy rainfall events. On the basis of these findings, it is recommended to improve satellite precipitation retrieval algorithms by incorporating topographical and local climatic factors into consideration.

A 24-h design storm for the Fort McMurray region

The time distribution of rainfall is an important input to rainfall-runoff modelling and has a significant effect on the resulting peak discharge. At many facilities in the oil sands region of Alberta, the Natural Resources Conservation Service (NRCS) Type II design storm has typically been used for design of surface water management facilities. The appropriateness of the Type II design storm and several other time distributions was assessed by comparing the modelled peak runoff from the design storms with the modelled runoff from historical 24-h rainfall events in five small typical catchments. Results indicate that the NRCS Type II produces highly conservative runoff peaks compared to actual rainfall events at Fort McMurray, and a new synthetic design storm developed during the study more closely parallels the peak runoff produced by actual rainfall events.

A comparison of de-noising methods for differential phase shift and associated rainfall estimation

Measured differential phase shift ФDP is known to be a noisy unstable polarimetric radar variable, such that the quality of ФDP data has direct impact on specific differential phase shift KDP estimation, and subsequently, the KDP-based rainfall estimation. Over the past decades, many ФDP de-noising methods have been developed; however, the de-noising effects in these methods and their impact on KDP-based rainfall estimation lack comprehensive comparative analysis. In this study, simulated noisy ФDP data were generated and de-noised by using several methods such as finite-impulse response (FIR), Kalman, wavelet, traditional mean, and median filters. The biases were compared between KDP from simulated and observed ФDP radial profiles after de-noising by these methods. The results suggest that the complicated FIR, Kalman, and wavelet methods have a better de-noising effect than the traditional methods. After ФDP was de-noised, the accuracy of the KDP-based rainfall estimation increased significantly based on the analysis of three actual rainfall events. The improvement in estimation was more obvious when KDP was estimated with ФDP de-noised by Kalman, FIR, and wavelet methods when the average rainfall was heavier than 5 mm h ≥1. However, the improved estimation was not significant when the precipitation intensity further increased to a rainfall rate beyond 10 mm h ≥1. The performance of wavelet analysis was found to be the most stable of these filters.

Sediment Wash-Off from an Impervious Urban Land Surface

AbstractSediment wash-off from impervious land is believed to be related to various factors including sediment buildup, rainfall intensity, rainfall energy, runoff rate, sediment sizes, and land surface characteristics. Most previous studies have used controlled rainfall in small areas to investigate mechanisms of sediment wash-off. This paper presents sediment wash-off results from an impervious urban surface under actual accumulation and rainfall events. Two temporal scales were considered when the sediment wash-off processes were investigated. On the event scale, sediment wash-off is strongly correlated with runoff volume. On the instantaneous scale, sediment wash-off from eight rainfall events was investigated using first-order and zero-order wash-off equations. Source-limited and transport-limited wash-off processes were discussed. The wash-off process on October 10, 2002, was linearly dependent on runoff and can be described using a first-order exponential equation. The wash-off processes on May 15,...