HS7.6

Did you know there are millions of rain observations from thousands of privately owned automated weather stations located throughout Britain (and beyond) held in a freely accessible online archive? Citizen Scientists are sharing detailed sub-daily weather observations, including from locations where other gauge data is not available, often in close to real-time. There is distinct clustering of rain gauges in British urban areas, and with an anticipated increase in convective storms resulting in localised pluvial flooding, such high-resolution data should not be ignored. The aims of this research are to assess data quality, investigate how access to the data can be made easier, and to explore how the data can be used to support improved flood risk assessment.

British rain observations are presented, spanning 10 years from more than 3000 unique citizen science weather stations via the Met Office WOW archive. These citizen science observations have the potential to fill gaps in the official monitoring network run by the Met Office and agencies responsible for flooding in Britain. Analysis indicates that if the official ground based rain gauge network was interpolated on a 5km grid there would be coverage for 36% of Britain, but if citizen science weather stations were included that figure increases to over 50%. A methodology to identify poor quality observations has been developed; the preliminary findings show that even where absolute values may be inaccurate, citizen science gauges can capture the pattern of extreme rainfall. Examples are shown from work in progress showing how combining citizen science observations with official rain data (radar and ground based gauges) can improve delineation of specific events that resulted in pluvial flooding.

How to cite: O'Hara, T., Parkin, G., Fowler, H., Lewis, E., McClean, F., and Brown, J.: Utilising Citizen Science Rain Data for Improved Rainfall Estimation in Urban Pluvial Flooding, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8734, https://doi.org/10.5194/egusphere-egu21-8734, 2021.

An operational, single-polarized X-band weather radar provides measurements in Hamburg’s city center for almost eight years. This weather radar operates at an elevation angle (~3.5°) with a high temporal (30 s), range (60 m), and sampling (1°) resolution resulting in a high information density within the 20 km scan radius. Studies on short time periods (several months) proofs the performance of this low-cost local area weather radar. For example, a case study on a tornado in a rain event demonstrates its refined resolution compared to the German nationwide C-band radars. Now, we aim for a eight-year precipitation climatology with 100 m resolution. This data set will enable reliable studies on urban extreme precipitation. This presentation will describe how we can infer a precipitation estimate based on multi-year weather radar observations in the urban area of Hamburg.

The single-polarization and small wavelength comes along with high resolution but at the same time high uncertainties. We address several sources of errors affecting the radar-based precipitation estimate, like the radar calibration, alignment, attenuation, noise, non-meteorologial echoes, and Z-R relation. The deployment of additional vertically pointing micro rain radars yields drop size distributions at relevant heights reducing errors effectively concerning the radar calibration and required statistical relations (k-Z and Z-R relation). We outline the performance of the correction methods for long time periods and discuss open issues and limitations.

With this high-quality and -resolution weather radar product, refined studies on the spatial and temporal scale of urban precipitation will be possible. This data set will be used for further hydrological research in an urban area within the project Sustainable Adaption Scenarios for Urban Areas – Water from Four Sides of the Cluster of Excellence Climate Climatic Change, and Society (CliCCS).

How to cite: Burgemeister, F., Clemens, M., and Ament, F.: Towards a multi-year urban precipitation climatology at 100 m scale using X-band radar observations, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1411, https://doi.org/10.5194/egusphere-egu21-1411, 2021.

Measuring urban precipitation adds extra difficulty to the already challenging task of quantitative precipitation estimation. Buildings form obstructions that can block ground-based precipitation radar signals, and the complex urban microclimate makes gauge measurements representative for only a very small area. Performing precipitation measurements in an urban setting thus benefits from using many different data sources, to capture the largest possible range of scales. As such, opportunistic sensing techniques are especially valuable for urban hydrometeorological research: the use of unconventional data sources to extract valuable data that can allow us to estimate urban precipitation. One of the more prominent data sources is the use of Commercial Microwave Links –CMLs – to measure rainfall, by making use of the signal attenuation between cell phone towers. This method of estimating rainfall has been mostly tested and applied in developed countries that already have reasonable coverage of conventional precipitation measurements. However, the most benefits are to be made in developing regions lacking such measurement networks. Only few studies address this, generally using relatively small datasets.

This research focuses on tropical CML rainfall estimation in Lagos, Nigeria. This African megacity has a dense network of CMLs and few official measurement stations, making it an interesting area to study the effectiveness of urban CML precipitation measurements in such a region. We employ the open-source R package RAINLINK to obtain 15-min rainfall maps based on data from a few thousand CMLs during the rainy season. We optimise the most important RAINLINK parameters by comparing to rain gauge data, considering local network and environmental conditions. In addition, disdrometer data from Nigeria or similar climates are used to compute the values of the physically-based coefficients relating specific attenuation to rainfall rate.

How to cite: Droste, A., Overeem, A., Priebe, J., Tricarico, D., Bogerd, L., Leijnse, H., and Uijlenhoet, R.: Measuring urban rainfall with a dense Commercial Microwave Link network in Lagos, Nigeria, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12325, https://doi.org/10.5194/egusphere-egu21-12325, 2021.

A novel proposal to create probabilistic attenuation nowcasting as a by-product from ensembles of rainfall forecasts is presented in this study. These attenuation nowcasts may eventually be used by mobile network operators to dynamically adjust their wireless network operations in advance and during heavy and extreme rainfall events. It may also facilitate mobile network operators to see a direct benefit of widely sharing its received power level data of their backhaul towers for 'opportunistic' rainfall estimation in real-time in urban areas becoming a clear win-win situation for telecom operators and hydrologists. It is proposed here that probabilistic attenuation forecasts can be derived from the ensembles of high-resolution forecast rainfall fields with lead times of 15 to 90 minutes generated from weather radar using the Short-Term Ensemble Prediction System (STEPS). The ensembles of rainfall predictions can be easily converted to attenuation for specific operating frequencies. This study used 109 microwave links ranging from 15 to 40 GHz to verify the results of this probabilistic attenuation forecast. Results suggest that the STEPS-based attenuation forecast was within the narrow span of the 90 percent confidence region for all microwave links tested, with up to 30-minute lead time, and was found to be skilful for lead times of up to 30-45 minutes. 

How to cite: Pudashine, J., Velasco-Forero, C., Curtis, M., Guyot, A., R.N. Pauwels, V., P. Walker, J., and Seed, A.: A win-win situation for telecom operators and hydrologists: A proof of concept of providing probabilistic attenuation nowcasting for the telecommunication networks, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13664, https://doi.org/10.5194/egusphere-egu21-13664, 2021.

This study investigated the wet deposition of particulate matter (PM) for six precipitation events at Daeyeon dong, Busan, South Korea, from February 2020 to July 2020. The concentration of PM₁₀ in the atmosphere was steadily measured before and after the precipitation. Rainwater samples were collected every 50 mL of each precipitation event using rainwater collecting devices and rainwater qualities (pH, electrical conductivity (EC), water-soluble ions (SO₄^2-, NO₃^-, NH₄⁺, Ca²⁺, etc.) were analyzed. For heavy rain events with strong rainfall intensities (>7.5 mm/h), the average PM₁₀ reduction efficiency reached more than 68%. For the relatively weak (<5 mm/h) rainfall intensities, the PM₁₀ reduction efficiencies were less than 40%. In all rainfall events, the average rainwater pH gradually increased over time from 4.3 to 5.0, while the average EC decreased from 81.9 to 12.1 µS/cm. The concentrations of all ions in the rainwater samples gradually decreased during precipitation. For heavy rain events, the acidity, EC, and concentrations of total water-soluble ions of initial rainwater samples were higher than those of later samples. This result indicates that the concentration of PM₁₀ in the atmosphere was reduced by wet deposition.

How to cite: Park, H. and Yang, M.: Variation of the concentration of particulate matter in the atmosphere and rainwater quality during precipitation at an urban site of southeast Korea, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3760, https://doi.org/10.5194/egusphere-egu21-3760, 2021.

Stochastic rainfall modelling is an increasingly popular technique used by the water and weather risk industries. It can be used to synthesise sufficiently long rainfall time series to support hydrological applications (such as sewer system design) or weather-related risk analysis (such as excess rainfall insurance product design). The Bartlett-Lewis (BL) rectangular pulse model is a type of stochastic model that represents rainfall using a Poisson cluster point process. It is calibrated with standard statistical properties of rainfall data (e.g. mean, coefficient of variation, skewness and auto-correlation and so on), but it can well preserve extreme statistics of rainfall at multiple timescales simultaneously. In addition, it is found to be less sensitive to observational data length than the existing rainfall frequency analysis methods based upon, for example, annual maxima time series, so it provides an alternative to rainfall extremes analysis when long rainfall datasets are not available. 

In this work, we would like to introduce an open source Python package for a BL model: pyBL, implemented based upon the state-of-the-art BL model developed in Onof and Wang (2020). In the pyBL package, the BL model is separated into three main modules. These are statistical properties calculation, BL model calibration and model sampling (i.e. simulation) modules. The statistical properties calculation module processes the input rainfall data and calculates its standard statistical properties at given timescales. The BL model calibration module conducts the model fitting based upon the re-derived BL equations given in Onof and Wang (2020). A numerical solver, based upon Dual Annealing optimization and Nelder-Mead local minimization techniques, is implemented to ensure the efficiency as well as to prevent from being drawn to local optima during the solving process. Finally, one can use the sampling module to generate stochastically rainfall time series at a given timescale and for any required data length, based upon a calibrated BL model.

The design of the pyBL is highly modularized, and the standard CSV data format is used for file exchange between modules. Users could easily incorporate given modules into their existing applications. In addition, a team, consisting of researchers from National Taiwan University and Imperial College London, will consistently implement the new breakthroughs in BL model to the package, so users will have access to the latest developments. The package is now undergoing the final quality check and will be available on Github (https://github.com/NTU-CompHydroMet-Lab/pyBL) in due course. 

How to cite: Wang, L.-P., Dai, T.-Y., He, Y.-T., Chou, C.-C., and Onof, C.: pyBL: An open source Python package for stochastic high-resolution rainfall modelling based upon a Bartlett Lewis Rectangular Pulse model, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8557, https://doi.org/10.5194/egusphere-egu21-8557, 2021.

Flooding is one of the most challenging weather-induced risks in urban areas, due both to the typically high exposures in terms of people, buildings, and infrastructures, and to the uncertainties lying in the modelling of the involved physical processes. The modelling of urban flooding is usually performed by means of different strategies in accordance with the specific purpose of the analysis, ranging from detailed simulations, requiring large modelling and computational efforts, and typically adopted for design purposes, to simplified evaluations, particularly feasible for scenario analyses, when a large number of simulations is required perturbing one or more input parameters.

According to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, intensity of precipitation events could be greatly impacted by the expected climate change primarily due to the increase in temperature, entailing an increase in the atmospheric moisture retention capability. However, the effect of climate change on the rainfall regime of local areas is not straightforward, but deeply depends on local features such as latitude, topography, distance from the coast. Over Europe, an ensemble of climate simulations coming from the application of different Regional Climate Models (RCMs) (able to perform a dynamical downscaling of General Circulation Models, GCMs, available at the global scale) is freely available within the EURO-CORDEX initiative, which is the current standard for climate change analysis over EU countries. The spatial resolution of EURO-CORDEX simulations (about 12km) is too coarse to be directly used in local impact analyses; in this case, bias corrections are usually performed using local rainfall observations, to adjust climate simulation results to the local rainfall regime. The availability of multiple climate projections coming from different Climate Simulation Chains (in other words, different RCM/GCM couplings) allows to quantify the uncertainty in climate modelling, that should be accounted for in impact analyses.

In the present work, an approach is proposed that aims to quantify the uncertainty caused by the use of an ensemble of climate projections on urban flood modelling, taking a limited area within the City of Naples (Italy) as test case. The specific purpose is that of understanding the resilience of the area with respect to any variation in rainfall intensity such as those possibly caused by climate change, building on 19 climate projections available within the EURO-CORDEX initiative and bias-corrected to make them suitable to be used for impact analyses at the local scale. The concept of resilience is expressed by a selection of indicators considered useful both in the framework of classical hazard analysis and for transport network, considered a strategic service for the test case. Urban flood modelling is undertaken by using two different numerical codes characterized by two different levels of complexity. In this way, it will be possible to draw conclusions about the computational costs that are actually needed, in terms of input data and resources, when integrating uncertainties due to climate projections in urban flood modelling for multi-purpose analyses.

How to cite: Padulano, R., Rianna, G., Costabile, P., Costanzo, C., Del Giudice, G., and Mercogliano, P.: Measuring urban resilience to flooding under climate change, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9705, https://doi.org/10.5194/egusphere-egu21-9705, 2021.

Most 1D-2D urban drainage models simplify the water exchange process between the ground surface and the underground sewer network. They assume that surface runoff enters directly into the sewer network without modelling the initial overland flow. Moreover, surface flooding is only seen as the result of sewer network overflows, neglecting that it can also occur due to limitations of street inlets draining capacity. In fact, it is a common practice to ignore street inlets and to assume that water exchange occurs exclusively through manholes, which in reality, can have water transfers (sewer overflow) only after the displacement of their covers. These simplifications, do not allow to determine which water exchange processes have a greater impact in the occurrence of surface flooding.

This study developed a more realistic model representation of the urban drainage system (including street inlets and initial overland flow) and carried out a thorough sensitivity analysis of the parameters controlling water exchange processes. A combination of Latin-Hypercube (LH) and One-factor-At-a-Time (OAT) sampling techniques were used to measure global and local sensitivities. Brederode neighborhood in Antwerp (Belgium), a flat area that frequently suffers from pluvial flooding, was used as a study case. Results show that the inclusion of street inlets reduces the calculated total surface flood volume in simulations with design storms ranging from low to high return periods (T5-T100). It was also found that parameters controlling surface drainage are the most sensitive, with the street inlets clogging coefficient obtaining the highest sensitivity index value. Parameters controlling reverse flow showed almost null sensitivity.

Given that the draining processes are most sensitive (particularly street inlets clogging) to the occurrence of surface flooding, urban drainage models should explicitly include manholes and street inlets in their configuration. Moreover, it is recommended to apply rainfall directly on 2D mesh elements representing streets and open areas (for runoff produced on rooftops, use sub-catchment units). In this way, models can account for initial overland flow and properly assess the street inlets' drainage capacity.

How to cite: González, A. and Willems, P.: Urban flood modelling and sensitivity analysis: water exchange between the ground surface and the underground sewer network, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1812, https://doi.org/10.5194/egusphere-egu21-1812, 2021.

The megacity of Lagos, Nigeria, is subject to recurrent severe flood events as a consequence of extreme rainfall. In addition, climate change might exacerbate this problem by increasing rainfall intensities. To study the hazard of pluvial flooding in urban areas, several complex hydraulic models exist with a high demand in terms of required input data, manual preprocessing, and computational power. However, for many regions in the world only insufficient local information is available. Moreover, the complexity of model setup prevents reproducible model initialisation and application. This conference contribution addresses these issues by an example application of the complex hydrodynamic model TELEMAC-2D for the city of Lagos. The complex initialisation procedure is simplified by the new package ‘telemac’ for the statistical environment R. A workflow will be presented that illustrates the functionality of the package and the use of publicly available information, such as free DEMs and Openstreetmap data to cope with the problem of insufficient local information. By further analysis and visualisation procedures along the workflow the increasing hazard of pluvial flooding for Lagos is shown. The workflow makes model initialisation, application, and the analysis of results reproducible and applicable to other regions with a relatively low need for manual user interventions and without additional software other than R and TELEMAC-2D.

How to cite: Pilz, T.: Towards reproducible pluvial flood simulation in urban areas with TELEMAC-2D, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5165, https://doi.org/10.5194/egusphere-egu21-5165, 2021.

In the face of rapid urban and population growth and with climate change altering precipitation patterns, urban water management is becoming increasingly demanding. Numerous software, tools and approaches to study urban water flow behaviour and model hydrological processes exist. However, the understanding of water movement in urban areas, especially during extreme events, and the physical principles behind them, as well as the interaction between the natural and the urban hydrological cycle is still incomplete. For decades, models suited for urban hydrological analysis greatly impacted the improvement of flood protection, public health and environmental protection, changing the way we look at urban water and stormwater management. In order to calculate accurate quantities of runoff in any rainfall/runoff model, information about urban sub-catchment characteristics plays an important role. Size, shape, topography, as well as land use influencing infiltration rates and evapotranspiration, are of great importance to calculate accurate runoff quantities on the urban scale. New implementations to reduce runoff towards the sewer system, such as decentralised stormwater techniques, increase the urgent need for accurate and high-resolution local/neighbourhood-scale information. Spatial and temporal developments require water management models to be connected with GIS (Geographical Information Systems). Initially not being developed to interact with each other, multiple approaches exist to combine GIS with water management models. Nevertheless, defining urban sub-catchments for rainfall-runoff modelling is often still performed manually using specific maps or using simple surface partitioning algorithms such as the Thiessen polygons. A significant disadvantage in generating urban sub-catchments manually is the fact that natural surface inclination is usually not considered, influencing the size and shape of the delineated sub-catchments. So far, only a few studies have devoted attention to improving the way urban sub-catchments are delineated and the information about their surface characteristics is generated. This study evaluates a GIS-based approach to automatically delineate urban sub-catchments accounting for the location of nodes (actual manholes or drain inlets) as sub-catchment outlets. In order to compare the influence of the sub-catchment delineation methods (1 to 3), we use (1) a digital surface model (DSM) and (2) a digital elevation model (DEM) to automatically delineate the urban sub-catchments and compare these two methods with each other as well as with (3) already manually derived sub-catchments of a specific case study. Furthermore, we compare hydraulic simulation results from the software SWMM with measured flow data to infer the most accurate sub-catchment delineation method.

How to cite: Back, Y., Funke, F., Bach, P. M., Leitao, J. P., Rauch, W., and Kleidorfer, M.: Comparing urban sub-catchment delineation approaches for dynamic hydrological modelling, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12004, https://doi.org/10.5194/egusphere-egu21-12004, 2021.

In this work a two-stage (rainfall nowcasting + flood prediction) analogue model for real-time urban flood forecasting is presented. The proposed approach accounts for the complexities of urban rainfall nowcasting while avoiding the expensive computational requirements of real-time urban flood forecasting.

The model has two consecutive stages:

• (1) Rainfall nowcasting: 0-6h lead time ensemble rainfall nowcasting is achieved by means of an analogue method, based on the assumption that similar climate condition will define similar patterns of temporal evolution of the rainfall. The framework uses the NORA analogue-based forecasting tool (Panziera et al., 2011), consisting of two layers. In the first layer, the 120 historical atmospheric (forcing) conditions most similar to the current atmospheric conditions are extracted, with the historical database consisting of ERA5 reanalysis data from the ECMWF and the current conditions derived from the US Global Forecasting System (GFS). In the second layer, twelve historical radar images most similar to the current one are extracted from amongst the historical radar images linked to the aforementioned 120 forcing analogues. Lastly, for each of the twelve analogues, the rainfall fields (at resolution of 1km/5min) observed after the present time are taken as one ensemble member. Note that principal component analysis (PCA) and uncorrelated multilinear PCA methods were tested for image feature extraction prior to applying the nearest neighbour technique for analogue selection.
• (2) Flood prediction: we predict flood extent using the high-resolution rainfall forecast from Stage 1, along with a database of pre-run flood maps at 1x1 km² solution from 157 catalogued historical flood events. A deterministic flood prediction is obtained by using the averaged response from the twelve flood maps associated to the twelve ensemble rainfall nowcasts, where for each gridded area the median value is adopted (assuming flood maps are equiprobabilistic). A probabilistic flood prediction is obtained by generating a quantile-based flood map. Note that the flood maps were generated through rolling ball-based mapping of the flood volumes predicted at each node of the InfoWorks ICM sewer model of the pilot area.

The Minworth catchment in the UK (~400 km²) was used to demonstrate the proposed model. Cross‑assessment was undertaken for each of 157 flooding events by leaving one event out from training in each iteration and using it for evaluation. With a focus on the spatial replication of flood/non-flood patterns, the predicted flood maps were converted to binary (flood/non-flood) maps. Quantitative assessment was undertaken by means of a contingency table. An average accuracy rate (i.e. proportion of correct predictions, out of all test events) of 71.4% was achieved, with individual accuracy rates ranging from 57.1% to 78.6%). Further testing is needed to confirm initial findings and flood mapping refinement will be pursued.

The proposed model is fast, easy and relatively inexpensive to operate, making it suitable for direct use by local authorities who often lack the expertise on and/or capabilities for flood modelling and forecasting.

References: Panziera et al. 2011. NORA–Nowcasting of Orographic Rainfall by means of Analogues. Quarterly Journal of the Royal Meteorological Society. 137, 2106-2123.

How to cite: Onof, C., Chen, Y., Wang, L.-P., Jones, A., and Ochoa Rodriguez, S.: A two-stage analogue model for real-time urban flood forecasting, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15645, https://doi.org/10.5194/egusphere-egu21-15645, 2021.

On the world’s fastest urbanizing continent, Africa, urban floods are a real and growing problem. Early warning is the first important step in flood risk management. This requires continuous and reliable precipitation measurements and forecasts, which are not always available in African cities.

In this study a nowcasting model based on Convolutional Neural Network (TrajGRU) was developed for short-term, 0-2 hours, precipitation forecast in Ghana, West Africa. The nowcasting model is trained on historical rainfall estimates derived from the MSG-SEVIRI instrument by the Nighttime Infrared Precipitation Estimation (NIPE) model. Input for the model are real-time NIPE MSG-SEVIRI estimates.

The Meteosat Second Generation (MSG) SEVIRI instrument provides high-resolution and short-latency data, covering Europa and Africa. Especially in areas without radar observations, MSG offers unique and relevant information for early warning with respect to fast occurring events such as urban flash floods. Its resolution allows for the retrieval of convective rainfall, often a cause of flash floods in tropical areas.

To assess the performance of the model, we compare our method to current state-of-the-art Lagrangian nowcasting system from the pySTEPS library applied to the NIPE-MSG-SEVIRI data.

The result is an operationally running model for nowcasting two hours ahead with 15 minutes temporal and approximately three kilometer in Ghana (rainsat.net). The method is readily applicable in other regions in Africa.

How to cite: Lugt, D., van Hoek, M., Meirink, J. F., and van der Kooij, E.: Nowcasting for urban flash floods in Africa: a machine-learning and satellite-observation based model, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-16002, https://doi.org/10.5194/egusphere-egu21-16002, 2021.

Over the past decades various types of permeable pavements have been implemented in different municipalities in the Netherlands in order to improve infiltration capacity in urban areas and therewith being able to better treat stormwater runoff. With initial promising results this adaptation measure seemed to be the solution for urban flooding due to extreme precipitation. However, in practice, foreseen infiltration capacities were usually not met, often due unknown reasons.

To better understand the functioning of permeable pavements in practice, we have studied - as part of the project Infiltrating Cities - over 100 existing permeable pavement installations in the Netherlands. At each location, infiltration capacity was tested through a full-scale infiltration testing procedure (flooded area about 40 m²) while conditional on-site factors were collected (location, age, type of permeable pavement, street-type, traffic density, vicinity of urban green, regular maintenance regime, etc.). By coupling this information we analyzed how these factors influence the infiltration capacity of permeable pavements in practice, e.g. through accelerated deterioration of infiltration capacity through time. In addition, we evaluated for a selected number of installations, how various types of maintenance may counteract this deterioration, hence improving the infiltration capacity of permeable pavements.

Most of he studied permeable pavements function, with an average infiltration capacity of 540 mm/hour, above Dutch and international standards. However, the observed variation in measured infiltration capacity is high (35 mm/hour – 5707 mm/hour) and cannot alone be explained by differences in age of the permeable pavement installations studied. Our analysis shows that also the deterioration-rate of the measured infiltration capacity, with an average of 74 mm/hour per year, varies substantially among installations, caused by factors like the vicinity of urban green, traffic density, and maintenance regime. The results have been compared to international studies finding similar conclusions about infiltration capacity and dominant factors, but little information is available of the effect of maintenance to recover the initial infiltration capacity. Evaluating the infiltration capacity after the application of various maintenance techniques shows us that applying the right maintenance regime to permeable pavements may improve infiltration capacity with an average of 380 percent. Especially in the case of under-performing permeable pavements this may be the key to improving the functioning and lifetime of permeable pavements in practice.  

Our results can be used to improve model representations of urban hydrological processes, give insights in potential adaptation strategies to deal with challenges related to (extreme) precipitation, and provide guidelines to city design in the light of climate change and rapid urbanization. Hence, various disciplines and user-groups can benefit from our outcomes: From the hydrological scientists aiming at improving the representation of urban hydrological processes in models order to better understand and predict how (extreme) precipitation may lead to urban flooding – now and in the future; to the urban water managers who are about to decide on the optimal strategy to deal with extreme precipitation and minimize urban flooding; and finally, to the urban designers that are developing resilient designs for future-proof cities.

How to cite: Veldkamp, T., Boogaard, F., de Graaf, R., and Kluck, J.: Understanding urban hydrology through measurements of infiltration capacity of permeable pavements under real-live circumstances, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-14547, https://doi.org/10.5194/egusphere-egu21-14547, 2021.

The Bioinfiltration Traffic Island (BTI) is a bioinfiltration rain garden that was retrofitted off an existing traffic island located at Villanova University, USA in 2001. Having been monitored since 2003, the BTI has quantitative hydrological data collected for the approximately two decades, making it a very valuable dataset for an in-depth analysis of the performance of the site.

The initial analysis comprises of a high-resolution analysis of rainfall event frequency along with resulting performance at the bioinfiltration rain garden. All rainfall events within the 15 years of collected data was discretized in to 2.5 mm, 2-hour bins. The binned rain events were then analyzed using a mass balance approach to understand how the different hydrological elements contribute to the ability of the site to treat incoming stormwater runoff. The second part of the analysis focuses on assessing the intensities of each of the recorded storms to understand its influence on the performance of the rain garden.

The third part of this analysis will comprise of studying the site’s ability to manage incoming runoff with the rain garden’s development over the fifteen years. The main focus will be to assess the performance of the site in the earlier stages and compare it to the performance seen in the latter stages with established vegetation. The binned rainfall events will be used to compare the performance of the BTI for storms with similar characteristics (similar precipitation amount and event duration) occurring in different stages of the timeline. The extensive dataset will additionally give insight to internal mechanisms such as evapotranspiration and infiltration occurring at the site and indicate how their contribution changes with the evolution of the site.

The overall analysis provides lessons into system components and aims to understand the interaction of the different hydrological elements within the rain garden. The objective is to use the findings in designing Green Infrastructure systems that can be optimized in their ability to manage incoming stormwater.

How to cite: Amur, A., Wadzuk, B., and Traver, R.: Analyzing the performance of a rain garden over 15 years, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12987, https://doi.org/10.5194/egusphere-egu21-12987, 2021.