Hydrodynamic Study of the Impact of Extreme Flooding Events on Wastewater Treatment Plants Considering Total Water Level

Extreme weather events such as hurricanes, severe storms, and flash floods have created havoc in New York (NY) state. They also have revealed the necessity of responding promptly to such weather events that threaten the life and property of New Yorkers. Hurricanes such as Sandy, Irene, and tropical storm Lee have increased the need for the preparation required to face the challenges of climate change. Several studies highlighted the catastrophic effect of Hurricane Sandy, both on human life and infrastructure (Abramson and Redlener 2012; Xian et al. 2015; Fahad et al. 2019). Hurricane Irene caused five deaths in North Carolina (Casey-Lockyer et al. 2013). At least a dozen people died from Tropical Storm Lee, which affected millions of people across the East Coast (Daniel 2011). In addition, these storms caused substantial damage to homes, businesses, transportation networks, and municipal infrastructure such as wastewater treatments and potable water treatment facilities. This damage left thousands of people without electricity. Another historical storm, Hurricane Ivan, produced a record 4.57-m (15-ft) storm surge on the coast of Florida, causing significant damage to the nearest wastewater treatment plant (WWTP), located in Pensacola Bay (Hummel et al. 2018). Undoubtedly, these extreme storm events, coupled with global warming, pose substantial threats to human safety, health, and hygiene, as well as to the critical infrastructure of human civilization (Uddin et al. 2019; Fahad et al. 2020; Sabrin et al. 2020). Despite the indispensable necessity of various infrastructure such as water, sanitation, transportation, and communication, there is an unmet need to overcome the lack of coordination among systems, the fragmented nature of vital information, and the interdependencies of natural and human-made systems. According to Comes and Van de Walle (2014), decision makers worldwide are leaning more toward improving resiliency and reducing vulnerability than eliminating the exposure to or likelihood of an extreme storm event as part of proactive resiliency measures.

It has been shown that climate change intensifies storm impacts and creates more damaging floods in coastal and low-lying areas. Starting with the dawn of the Industrial Age in the nineteenth century, the rate of human-induced sea level rise has been accelerating, and during the twentieth century, sea level rose roughly 0.18 m ($\sim 7\mathrm{in.}$) (USEPA 2020; Peteet et al. 2018; Lin and Shullman 2017). According to the recent climate projections, by the end of the 21st century, the temperature will increase by 4.8°C, with a sea level rise of as much as 0.82 m (IPCC 2014; Fahad et al. 2018).

New York City (NYC) is influenced by hurricanes, tropical storms, and nor’easters. Nor’easters can occur throughout the year, but are most frequent and strong between September and April. Coastal areas experience tropical storms between July and October, which produce high wind speeds and storm surges. The chances of coastal inundation increase when the storm surges combine with a high astronomical tide. Studies have shown that extreme spring tide in coastal NY could be 2 times higher ($\sim 2.68\mathrm{m}$) than the mean range of tide, 1.37 m (Horton et al. 2015; Kinney et al. 2015).

Hurricane Sandy was one of the most catastrophic and deadliest storms for coastal New York, generating $\sim \$68\mathrm{billion}$ in damage (Blake et al. 2013). Sandy hit NY Harbor on October 29, 2012, during a high tide in the Atlantic Ocean and impacted NYC tremendously. According to the report of the New York City Panel on Climate Change (NPCC), a flood elevation of 4.3 m (14.1 ft) was recorded at the Battery tide gauge during Hurricane Sandy (Horton et al. 2015). Hurricane Sandy also highlighted vulnerabilities in some of the city’s infrastructure, utilities, healthcare, telecommunication, transportation, and water and wastewater infrastructure (Blake et al. 2013; Tomiczek et al. 2017). Even years after Sandy, coastal flooding remains a significant concern for NYC because nearly $50\mathrm{million}{\mathrm{m}}^2$ of the built environment and 400,000 residents are within the current 100-year floodplain (Garner et al. 2018). Failure of near-coast WWTPs would pose an enormous environmental and economic impact on the people living near the coast.

NYC has one of the largest and most complex water supply and treatment systems, containing 14 WWTPs and 96 pumping stations that convey stormwater and wastewater. However, most WWTP systems are considered inadequate due to the increased intensity and frequency of tidal flooding, sea level rise (SLR), and storm surges (Allen et al. 2019; Karamouz et al. 2016a, b). Wastewater infrastructure plays a critical role in the urban built environment by providing the safe and efficient conveyance and treatment of sewage to protect human health and the environment. Most WWTPs in the US are located at low elevations near the coast to minimize development costs. These treatment plants can use gravity-fed systems to facilitate the conveyance of wastewater flow, thereby reducing the need for pumping stations. In addition, coastal locations allow for the efficient discharge of treated effluent to neighboring water bodies. In extreme weather conditions such as flooding, the performance of a WWTP might be decreased due to a sudden increase in quantity and deterioration of quality of influent and to damaged assets (Sweetapple et al. 2017). Decrease in elevation head during flooding causes failure of the gravity-fed collection system in WWTPs. Insufficient head differential for flow increases the chance of the discharge of untreated raw sewage to nearby water bodies. Loss of WWTP functionality also leads to halting the conversion process of turning wastewater into a reusable water source, causing severe issues for the urban population.

Because of Sandy, the sewage overflow in NYC was 5.2 billion gal. total, which included 6 spills of over 100 million gal. and 28 spills of 1 million gal. or more. New York City Department of Environmental (NYCDEP) reported 560 million gal. of untreated combined sewage discharged and 800 million gal. of partially treated combined sewage (SIRR 2012). According to NYCDEP, by the 2050s all 14 WWTPs will have some equipment below base flood elevation (BFE). At present, 37 of 96 pumping stations are in the 100-year floodplain, as indicated by FEMA (2013) preliminary working maps, and by the 2050s, the number is expected to increase to 58 of 96 (USEPA 2020). Evidence has showed that most overflow incidents were due to the loss of electrical power at critical equipment of the WWTPs, and 10 of the 14 wastewater treatment plants either were damaged or lost power during Hurricane Sandy. Three of the plants remained nonoperational for some time. Although the vulnerability of WWTPs historically has been highlighted numerous times, there is a tremendous lack of systematic understanding of the causes of, the severity of the damage from, and the economic consequences of flooding and extreme storm events for the WWTPs in NYC.

This study examined the causes of failures in wastewater treatment systems during extreme weather events through inventorying, mapping, and reviewing the performance of sewage treatment plants. The study implemented two hydrodynamic models, namely Sea, Lake, and Overland Surges from Hurricanes (SLOSH) and Hazus, to account for the flooding due to storm surge and extreme rainfall events. An in-depth analysis of the cost of damage due to severe storm events specifically for the WWTPs also was conducted using Hazus. In addition, we provided site-specific guidance to make the infrastructure resilient to extreme natural events. Climate change, sea level rise, and extreme natural events such as hurricanes are significant factors that cause flooding.

Modern civilization and urban infrastructures are endangered due to the combined effect of climate change and extreme storm events (Gebre and Ludwig 2015; Saiful Islam et al. 2018). Because of rising sea levels combined with extreme storm events, critical nearshore infrastructures such as WWTPs have been deemed likely to experience permanent inundation, forcing the costly retrofitting or relocation of existing facilities (Hummel et al. 2018). Since 1950, the sea level around Butler Township in New York has risen by nearly 0.23 m (9 in.); in the last decade the rate of increase has accelerated, and sea level now is rising by 0.025 m every 7–8 years. Studies have projected that NYC areas would be submerged under several meters of water due to the combination of sea level rise and intense storms such as Super Storm Sandy (Hatzikyriakou et al. 2016; Lu et al. 2019; CCPS 2020; Rashid et al. 2018). However, until now, all these events have been studied in singular form, and the compound effect has been ignored (Olyaei et al. 2018; Zouboulis and Tolkou 2015; Hummel et al. 2018; Cao et al. 2020). Specifically, in recent years, the primary damage has come from cascading events rather than from a single decisive factor. The critical factor that needs to be considered is what will happen if a hurricane surge is combined with multiple days of rain, sea level rise, and high tide during a significant event. A medium-sized storm such as Hurricane Sandy, with $22.35{\mathrm{ms}}^{-1}$ ($50\mathrm{mi}/\mathrm{h}$) wind speed, would have a devastating impact on New York and New Jersey, and Hurricane Harvey would destroy Texas and Louisiana, completely flooding Houston for days after 1,524 mm (60-in.) rainfalls in only 3 days, which is more than Texas normally receives in 1 year. Storm surge is only one factor contributing to water level rise along the coast during a hurricane. Other factors, including tides, waves, freshwater input, the central pressure of the hurricane, forward storm speed, the shape of the coastline, and sea level rise contribute to the storm surge height (Orton et al. 2012; Soomere and Pindsoo 2016; Serafin et al. 2017), and need to be considered for accurate prediction of water levels. This work suggests using the resulting impact of the combined factors, the total water level (TWL) (Fig. 1). Components of TWL can be defined as the projected mean sea level rise (PMSL), freshwater input (FI), higher high-water mark (HHW), storm surge (SS), wave setup (WS), and wave uprush (WU). The following equation describes the standard formulation for TWL:

The study utilized the linear summation of the TWL constituents to simulate the extreme flooding events for the WWTPs. Several studies have adopted similar methodology to simulate the extreme total water levels. For example, Serafin and Ruggiero (2014) simulated the elevation of the TWL considering linear summation of four components to explain the total contribution from mean sea level, astronomical tides, wave runup, and nontidal residuals. Melo de Almeida et al. (2018) determined the contributions of mean sea level rise to coastal flooding for southeastern Vietnam using a similar linear approach to quantify the total water level. Similarly, Almar et al. (2019) studied the influence of climate variability on tropical cyclones using the linear concept of TWL. Although the nonlinear assumptions still are valid for the model configuration of SLOSH and Hazus, the final flood maps were utilized as a linear summation for the constituents of TWL. Therefore, this study used a similar approach to quantify the total water level. The proposed TWL approach will develop the systematic understanding required to determine the damage and economic consequences of floods and extreme storms on WWTPs.

The data and methods used in the modeling process seriously affect the loss estimate’s uncertainty (Tate et al. 2015). Choosing the right research tool is extremely important. The primary modeling tools used for this work were SLOSH and Hazus-MH 4.2, which are predominantly GIS-based natural disaster loss estimation software packages (Scawthorn et al. 2006; Ding et al. 2008; Do et al. 2020). In this research work, Hazus-MH 4.2 was used to estimate the flood depths and elevations for coastal and riverine flooding for 10-, 25-, 50-, 100-, and 500-year flooding events. The main purpose of using two different models (i.e., SLOSH and Hazus) was to create the flooding scenarios for different types of extreme storm events. For example, flood maps generated from SLOSH modeling provided the risk of flooding based on different hurricane categories, whereas Hazus provided riverine and coastal flooding based on rainfall-induced flooding and return periods. Because SLOSH is unable to provide the flooding due to river flow or extreme rainfall, it was necessary to simulate those effects in Hazus to obtain a complete picture of the vulnerability of WWTPs due to TWL.

To consider the effect of hurricane storm surge from different hurricane categories, outputs from SLOSH were utilized. Historical hurricane events such as Hurricanes Sandy and Irene were considered for verification and comparison purposes using data sets provided by FEMA. FEMA provided data sets of Hurricane Sandy (FEMA 2013). These data sets are in raster format. In addition, particular focus was given to two WWTPs of NYC, Hunt’s Point and Rockaway WWTPs. We utilized the concept of TWL to assess the maximum probable effect of extreme storm surge on critical infrastructure such as WWTPs. Fig. 2 shows the general workflow to derive the inundation component for each of the factors associated with TWL. First, flooding due to hurricanes was derived from the SLOSH and topographic information. The contribution of coastal flooding for different return periods was modeled in Hazus-MH. Subsequently, the percentage damage due to flooding was calculated using the depth-damage function from the Hazus-MH model. The coastal and riverine flooding was considered by utilizing different categories of hurricanes in SLOSH and return period–based flooding in Hazus. Lastly, the contribution of climate change–induced sea level rise was incorporated as inundation maps considering 0.61 m (2 ft) of SLR for the 2050s.

This section describes coastal and riverine flooding modeling in the Hazus-MH flood model based on multiple return periods for the study area (i.e., NYC). The flood depth grids were developed based on high-resolution topographic information using the USGS 10-m digital elevation model (DEM). The floodwater depth was estimated based on the stream network generated by Hazus, and associated hydraulic analysis was carried out based on the topographic data provided (Gutenson et al. 2015). The importance of using a high-resolution DEM was highlighted by Banks et al. (2015), who showed that a higher-resolution DEM produces better damage estimation. The exposure to flooding for the WWTPs was identified based on 10-, 50-, and 100-year return period flooding, and the potential economic losses were calculated based on the depth damage function integrated into Hazus. Loss estimation through Hazus-MH demonstrates that the latest scientific and engineering knowledge can provide a reasonable basis for formulating mitigation measures and emergency preparedness, response, and recovery plans and policies, and ultimately can provide a basis for government decisions at all levels. The deep-damage function within Hazus is a relationship diagram between the flood depth and the percentage of destruction drawn for various structures and uses. We used the default database at the census tract level that contains information regarding the general building stocks in terms of building footprint and counts, essential services, including the WWTPs, conveyance system, and utility services. The inventory data were aggregated at a census tract level as defined by US census data of 2010. After the affected structures were identified—in this case, the WWTPs—flood damage functions were utilized in Hazus-MH to relate the height of floodwater to the corresponding amount of damage in terms of the total replacement cost. These depth–damage curves were developed by different sources, including the Federal Insurance Administration (FIA), the USACE, and the Institute for Water Resources (IWR), depending on the study area (Scawthorn et al. 2006). For the loss estimation, depth–damage curves are chosen by Hazus-MH from the library based on the type and its content. The wind force during hurricanes affects light-framed structures, whereas the different components of WWTPs usually are built of noncombustible materials such as concrete, steel, and brick. Therefore, this study considered the damage due to wind force and debris to be negligible for the case of WWTPs, and utilized the flood depth–damage functions within the Hazus model to calculate the percentage damage due to flooding and associated economic loss.

This study utilized the Sea, Lake, and Overland Surges from Hurricanes (SLOSH)-derived storm surge products to identify the flood inundation due to storm surges. SLOSH was developed by the National Weather Service to provide storm surge guidance in a variety of ways (Glahn et al. 2009; Forbes et al. 2014). The SLOSH modeling method can estimate fluctuations and can be deterministic (solving physical equations), stochastic (statistical processes), or compound (maximum water seal) according to the Saffir–Simpson hurricane intensity scale (Schott et al. 2012). SLOSH provides two composite products: the maximum envelopes of water (MEOW) and maximum of the MEOWs (MOM). It prepares a set of highest surge values at each grid location to target a given storm category, forward speed, and direction of movement. MOM is the synthesis of the maximum storm surge height of all simulated hurricanes in each category. Each basin generally corresponds to 5 MOMs (one for each storm category) (Lin and Shullman 2017). The MOMs contain the surge height values for Category 1–4 storms that make landfall at high tide. The National Hurricane Center obtained related data that include the surge height grids for the NY3 basin as gridded vector data. Using the USGS 10-m elevation data, the inundation depth was calculated based on geospatial analysis from the gridded storm surge height data. Subsequently, the flood depth information for the WWTPs was extracted from the inundation surface.

We used the FEMA Modeling Task Force (MOTF)-derived flood inundation products for Hurricanes Sandy and Irene as the observed data to assess the flooding of WWTPs. The FEMA MOTF specializes in modeling and risk analysis. To obtain loss estimates, the FEMA Modeling Task Force uses remote sensing analysts to examine detailed satellite and aerial images after a significant disaster, mainly for recovery planning and coordination purposes (Griffith et al. 2015). To provide the best estimate of the impact and debris before, during, and after an incident, it aggregates hazard and modeling information from various sources (including other federal agencies, universities, national laboratories, and state and local agencies). The MOTF integrates the information observed throughout the disaster into the ground truth (FEMA 2013). For Hurricane Sandy, these data sets were created based on the field-tested high-water mark (HWM) and storm surge sensor data provided by the USGS in this research. HWM and surge sensor data were used to interpolate the water surface elevation and establish a depth grid and surge boundary by state; these values were subtracted from the best available digital elevation model.

Affected WWTPs and associated inundation levels were determined using flood mapping analysis from two historical hurricane events (i.e., Hurricanes Sandy and Irene) for NYC.

Fig. 3 shows the MOTF inundation map of two hurricanes, Irene and Sandy, in NYC. Fig. 3(a) is the MOTF Hurricane Sandy inundation map. Most near-coast WWTPs were inundated during significant storm events with various water levels. There was a greater extent of inundation and an associated increased inundation level during Hurricane Sandy [Fig. 3(a)] than during Hurricane Irene [Fig. 3(b)]. Significantly, the WWTPs located near the east coast were affected more than those further inland by historical events. Analysis using observed inundation from historical hurricanes provides only part of the information for vulnerability assessment, rather than the full spectrum of damage considering every category of storms (i.e., Category 1, Category 2, Category 3, and so forth). The flood maps in Figs. 3(c–f) show the impact of all four categories of storms on the WWTPs of NYC. Although the Saffir–Simpson hurricane scale has five hurricane categories (Categories 1–5), the National Oceanic and Atmospheric Administration (NOAA) SLOSH model provides SLOSH maps for only four categories of hurricanes (Categories 1–4). Results showed an increase in WWTPs affected as the intensity of hurricane increased from Category 1 to Category 4. Except for Category 1 and 2 hurricanes, every WWTP located in NYC was severely inundated during extreme storm events, and Category 3 and Category 4 storms were the most destructive. To see the long-term resiliency plan for both the WWTPs, the average of the mean flood depth differences between different categories of the storm was calculated. The maximum value for Hunt’s Point WWTP was 1.53 m (5.017 ft), and for Rockaway WWTP it was 1.84 m (6.06 ft). This reveals that Rockaway WWTP is more prone to flooding than the Hunt’s Point WWTP, and the reason could be that Hunt’s Point is inland, whereas Rockaway WWTP is exposed to the Atlantic Ocean and Jamaica Bay.

Fig. 4 depicts the effect of different flooding events based on return periods for different WWTPs of NYC, and shows the flood depths and flood elevations as simulated by Hazus. The flood depth for a 100-year coastal flood event at Hunt’s Point WWTP was 1.94 m (6.372 ft), whereas at Rockaway WWTP, the flood depth for the same event was 2.84 m (9.333 ft). Likewise, SLOSH and Hazus-MH modeling indicated that both plants are expected to be affected by floods in all three return periods. Results of both models showed that Rockaway WWTP has a greater flood depth than Hunt’s Point WWTP. On the other hand, Hunt’s Point WWTP has little riverine flooding effect, which was negligible compared with the 100-year coastal flooding, whereas Rockaway WWTP was not affected by riverine flooding. Therefore, per Hazus-MH analysis, both WWTPs need a resiliency plan, and Rockaway WWTP needs more protection than Hunt’s Point WWTP in the long run. Furthermore, the hazard type also can be both riverine and coastal, considering 100-year and 500-year floods as examples. In addition, 8 and 10 of 14WWTPs severely flooded by 100-year and 500-year flooding events, respectively, according to the Hazus-MH analysis (Fig. 4). Because civil infrastructures are designed and constructed considering 25-year, 50-year, or 100-year events, it is essential to identify the associated damage level and economic loss before designing critical infrastructures such as WWTPs. Table 1 highlights the affected WWTPs and various property damage levels according to Hazus-MH.

Two WWTP suffered $\sim 40\%$ damage due to a 100-year flood event, and the total damage was $126,561,896 considering all the affected WWTPs (Table 1).

Using the data sets of the FEMA MOTF for Hurricane Sandy, the inundation maps were used to extract the flood depths and flood elevations for the two most vulnerable locations (i.e., Hunt’s point and Rockaway WWTPs). The maximum flood depth was 5.25 m (17.213 ft) and 4.36 m (14.312 ft) for Hunt’s point and Rockaway WWTPs, respectively. Overall, the average depth of inundation was between 0.38 and 1 m (1.256–3.278 ft).

Fig. 5 shows the Hazus 100-year and the Flood Insurance Rate Map (FIRM) 100-year maps. The comparative study showed that flood elevation estimation from both maps was very close, but the flood coverage was different: Hazus 100-year maps indicated more flooded areas. The difference of the DEMs used could cause this when generating flood depth grid, and it probably was because FIRM uses DEMs differently from the USGS 10-m DEM used by Hazus-MH. Because DEMs are created to represent the ground elevation, different DEMs have different ground elevation values. Thus, given the same flood elevation, different ground elevations will result in different flood depths.

In this study, considering Hunt’s Point WWTP as an example, ground elevations at locations which are covered by the Hazus flood but not by the FIRM flood can be speculated to be slightly lower in the USGS 10-m DEM. At Rockaway, the flood elevation estimated using the Hazus map was 17.57% greater than the FIRM map’s flood elevation, but flood coverage remained the same and the facility was flooded completely. Because the FEMA FIRM 100-year flood map also estimates a 100-year flood event in NYC, it is necessary to compare the results from it and from a Hazus 100-year coastal flood event. The complete comparison of Hazus and FIRM 100-year flood depths for all 14 WWTPs in NYC is presented in Table 2. Half of the WWTPs had smaller Hazus-computed flood depths than FEMA FIRM values, with an average difference of $-7.32\%$, and, the other half had an average difference of 11.77%.

Three WWTPs that had the most significant differences between Hazus 100-year flood elevation and the Advisory Base Flood Elevation (ABFE) 100-year flood elevation were Owls Head WWTP (21.27%), Oakwood Beach WWTP (20.67%), and Rockaway WWTP (17.57%) (Table 2). Unlike the other 11 WWTPs, these 3 WWTPs have a direct connection and exposure to the ocean. The still water elevation from the flood insurance study (FIS) report was 3.47 m for Owls Head WWTP, 3.65 m for Oakwood Beach WWTP, and 3.29 m for Rockaway WWTP (FIS-FEMA 2013). Considering the waves and wave splash due to the exposure to the ocean, the WWTPs have greater chance of flooding with higher level of inundation than the FEMA FIRM values.

Fig. 6 shows the level of flooding based on the return period simulated in Hazus. The two most vulnerable WWTPs, namely Hunt’s Point and Rockaway, were selected to showcase the results. For both WWTPs, the inundation level increased with the increase in return period, and Hunt’s Point was affected more than Rockaway. For relatively minor storms (i.e., 10-year and 50-year), the extent of inundation was much more significant for Hunt’s Point than for Rockaway [Fig. 6(a)]. The inundation maps also show that the effect of storm surge was much more substantial on the southern and eastern part of the plants, which are closer to the coasts [Fig. 6(a)].

Rockaway WWTP was more affected by riverine flooding than by coastal storm surge [Fig. 6(b)]. The flood depth and extent were much lower at Rockaway WWTP than at Hunt’s Point except for those from the 100-year storm events. Similar to Hunt’s Point WWTP, the maximum flood depth estimated for Rockaway WWTP for the 10-year event was 1.83 m, which increased to 2.13, 2.74, 4.26, and 5.18 m fir the 25-, 50-, 100-, and 500-year events, respectively (Fig. 6). The northeast corner of the map was susceptible to the maximum flood depth.

Hurricane Sandy was a Category 3 hurricane in the ocean. However, during landfall, it was only a tropical storm, with a wind speed of $22.35{\mathrm{ms}}^{-1}$ ($50\mathrm{mi}/\mathrm{h}$), which combined with other local cascading events and caused havoc on the East Coast of the US. Therefore, we introduce the concept of total water level to illustrate how the impact of a given storm event can be multiplied when it is combined with other concurrent synergetic factors. TWL analysis was conducted considering the two most probable hurricane landfall events (i.e., Categories 1 and 2) for NYC. For each hurricane category, we added the inundation depth from the most frequently occurring 100-year rainfall–induced flood depth and a sea level rise of 0.61 m (2 ft), which is the projected sea level rise for NYC in the 2050s. The sea level rise projection values were adjusted to reflect the long-term high tide or high high-water mark (HHWM) scenario for the best possible assessment. The results are shown in Fig. 7.

Figs. 7(a and b) show the spatial distribution of the inundation depth for NYC considering two pathways for TWL. Both the flood depth and the flood extent increased by a great extent for Category 2 hurricanes compared with Category 1. The inundation depth for the WWTPs was analyzed considering the footprint of the whole system, rather than a single-point analysis. The average depth of inundation for the two TWL cases is highlighted in Fig. 7(c). The compound effect of extreme storm events such as a hurricane with sea level rise and rainfall events will cause devastating flooding to critical infrastructure such as WWTPs. The two vulnerable WWTPs in NYC, Rockaway and Hunt’s Point, were subjected to an additional 2.74–4.27 m (9–14 ft) of flooding.

This research work was conducted to evaluate the impact of various types of extreme storm events on the WWTPs of NYC. This comprehensive study compared flood depth and extent caused by hurricanes, different return periods of flooding, and historical extreme storm events. The study proposes a new concept to identify the risk of flooding considering the total water level effect. The components for extreme storm events (i.e., hurricanes) were quantified using the SLOSH-derived inundation maps for different categories of the hurricane. The study also simulated the effect of coastal and riverine flooding due to extreme rainfall events using Hazus-MH. Associated damage levels were determined based on the depth–damage functions integrated into the Hazus model. The bathtub model determined sea level rise due to climate change to develop the inundation surface. We utilized various geospatial techniques to aggregate all the different flood inundation products to simulate the final flood maps considering the TWL. Associated depth of flooding also was determined for the WWTPs within New York City.

In addition, for verification and comparison purposes, the impact of Hurricane Sandy was considered using the data sets provided by the FEMA MOTF as an observed scenario. Results showed that the WWTPs close to the coastline, such as Rockaway WWTP, are more vulnerable to flooding than those situated inland. The analysis concluded that, for a specific WWTP, the depth of flooding can be different depending on the type of extreme event considered, and it can increase significantly when TWL is used to consider two or more simultaneous events. Thus, for design, maintenance, and long-term planning, it is essential to look at the overall vulnerability from a holistic perspective considering the combined effect of extreme storm events, sea level rise, rainfall, and high tide through the concept of total water level. Analysis showed the disastrous consequences of the compounding effect when considering all the possible factors contributing to flooding. All the concurrent events need to be considered for the accurate prediction of water levels. Otherwise, the projected flooding and damage amount will be significantly underestimated, as has been the case during recent years.

All data, models, or codes that support the findings of this study are available from the corresponding author upon reasonable request. This includes raw data, analysis results, figures, tables and codes that was generated as part of the study.

The New York State Resiliency Institute for Storms and Emergencies (NYS-RISE) was founded to address the challenges in the State. This work has been funded by a research project that the State of New York supports through NYS-RISE.