1. Introduction
Unexpected flash flooding is one of the most devastating natural hazards, causing economic losses and fatalities (e.g., Glancy and Harmsen 1975; Randerson 1976; Ogden et al. 2000; Ashley and Ashley 2008). The National Oceanic and Atmospheric Administration (NOAA) National Weather Service (NWS) recorded 1075 flash-flood-related fatalities across the United States from 1996 to 2014 (Gourley et al. 2013). Urbanized watersheds are more prone to flash flooding compared to natural, undeveloped watersheds because of the increased hydrologic efficiency stemming from impervious surfaces, compacted soils, insufficient drainage capacity, and channel lining (e.g., Leopold 1968; Smith et al. 2005a,b; Wright et al. 2012; Hodgkins et al. 2019; Zhu et al. 2019). Furthermore, rapid land development, aging and frequently undersized drainage conveyance systems, and often uncoordinated management at watershed scales have exacerbated the social and economic impacts of urban flash flooding across the United States. For example, multiple recent billion-dollar weather disasters are associated with extreme storms over urban areas, such as the 2014 Detroit, 2017 Houston, and 2018 Maryland floods (NCEI 2022).
Because of the impact of flash floods on life and property, it is important that they are adequately predicted. However, this task remains nontrivial due to the uncertainties in numerical weather predictions and error propagation through hydrologic and hydraulic processes (e.g., Collier 2007; Hapuarachchi et al. 2011). In the context of climate warming, the Clausius–Clapeyron relation indicates that rainfall extremes will increase at a rate of ∼7% °C−1 warming and >7% increases are expected in short-duration rainfall intensities in some regions stemming from positive feedbacks in convective dynamics (Payne et al. 2020; Fowler et al. 2021). More people are also projected to live in urban areas in the future because of urban development, population growth, and rural-to-urban migration (e.g., Mazumdar 1987; Lucas 2004). Therefore, the frequency and severity of urban flash floods are projected to increase substantially (e.g., Prein et al. 2017; Fowler et al. 2021; Li et al. 2022).
Previous studies have examined flash floods from both the synoptic and mesoscale perspectives. Maddox et al. (1979, 1980) found that exceedingly rare and strong flash flooding in the southwestern United States is associated with the North American monsoon (NAM)-related moisture surge and 500-hPa shortwave troughs moving along a longwave ridge. Flash-flood-producing storms typically occur ahead of shortwave troughs where the atmosphere is moist and unstable (e.g., the September 1974 Eldorado Canyon Flood; Maddox et al. 1980). Focusing on Arizona during the NAM season (15 June–30 September), Maddox et al. (1995) and Mazon et al. (2016) identified three synoptic-scale patterns of 500-hPa geopotential height that are associated with a great number of thunderstorms. These three patterns reflect the movement of a subtropical ridge and can be described as shifting northeastward into the central United States (Type I), shifting northwestward into Nevada and Utah (Type II), and receding southward (Type III).
The advent of Weather Surveillance Radar-1988 Doppler (WSR-88D; Heiss et al. 1990; Fulton et al. 1998) radars across the United States and their continuous improvement (e.g., dual polarization) has facilitated storm-scale analyses of flash flooding systems. By examining the 3D radar reflectivity and cloud-to-ground (CG) lightning flash rate, Smith et al. (2005a) showed that the 13 June 2003 extreme rainfall event in Baltimore, Maryland, resulted from a collapsing thunderstorm cell. Yang et al. (2019) analyzed a catalog of 102 storm events in central Arizona during the NAM season and argued that a broad spectrum of convective intensities can yield large hourly rainfall rates (>30 mm h−1). Smith et al. (2019) examined the devastating 14 September 2015 storm in Hildale, Utah, identifying melting hail, which enhanced downdrafts, as the principal agent for extreme convective intensity and rainfall rates. In addition, a few studies based on radar reflectivity and storm-tracking algorithms have showed that storm properties—including the orientation, shape, and motion of storm cells—play key roles in flash flooding in small- and medium-sized (<1000 km2) watersheds in arid to semiarid regions where antecedent soil moisture is typically dry (e.g., Davis 2001; Morin et al. 2006; Peleg and Morin 2012).
This study focuses on the arid 4084-km2 Las Vegas Wash (LVW) watershed in southern Nevada. The LVW watershed encompasses the extremely urbanized Las Vegas metropolitan area with a population of over two million (U.S. Census Bureau 2020). Despite its arid climate, the LVW watershed is prone to seasonal flash flooding due to large-scale atmospheric circulations, orographically enhanced heavy rainfall, relatively impermeable desert pavement, and rapid urbanization (Yu et al. 2023). During the cool season (defined as October–March), inland-penetrating atmospheric rivers (ARs) transport abundant moisture from the Pacific Southwest northeastward to the interior land, which may result in heavy rainfall over the southwestern United States, including LVW watershed (e.g., Rutz et al. 2014, 2015; Hatchett et al. 2017; Albano et al. 2020; Rivera et al. 2014). During the warm season (defined as April–September), convective storms are the main drivers for flash flooding in the watershed (e.g., Mazon et al. 2016; Yang et al. 2019; Smith et al. 2019).
To the best of our knowledge, previous studies on flash floods in the southwestern United States have mainly concentrated on coastal California in the cool season and on the NAM region (e.g., southern Arizona and New Mexico) in the warm season (e.g., Neiman et al. 2008; Dettinger et al. 2011; Oakley et al. 2017; Yang et al. 2019; Hatchett et al. 2020; Barlow et al. 2019). However, watersheds in the interior southwestern United States, such as the LVW watershed, may also be susceptible to seasonal flash floods. Therefore, the central objective of this study is to investigate the characteristics of storms that produce flash flooding for the LVW watershed in both warm and cool seasons.
We focused on storm events during the 1996–2021 period, for which a dense network of rain gauges and radar from the WSR-88D in Las Vegas (KESX) are available. First, we identified a relatively large sample of flash flood events and investigated their relationship to synoptic features, as well as several other meteorological and climatological variables. Second, we performed storm-scale analyses to understand both the convective intensity and motion of flash-flood-producing storms. Convective intensity is characterized as the maximum reflectivity and echo-top height of storms, as well as the CG lightning flash rate. Storm motion is examined from a Lagrangian perspective using 3D radar reflectivity and a storm-tracking algorithm. Lagrangian analyses of storms accommodate understanding the spatiotemporal structure of storms and their interactions with the watershed in generating and propagating flood waves. Finally, we discuss orographic and warming effects on the flash-flood-producing storms for the LVW watershed.
2. The urbanizing Las Vegas Wash (LVW) watershed
The arid 4084-km2 LVW watershed in southern Nevada is part of the Colorado River basin (Fig. 1). The elevation of the LVW watershed ranges from >2500 m above mean sea level (MSL) over the Spring Mountains and Sheep Range in the west and north, respectively, to lower than 700 m MSL over the alluvial valley in the southeast (Fig. 1). Several smaller mountains (<1000 m MSL) are located to the south and east of the valley. The ephemeral LVW channel crosses the valley floor from northwest to southeast before terminating in Lake Mead (Fig. 1). The native soil type is primarily fine sandy loam. The older geomorphic alluvial surfaces closer to the mountain fronts may be armored by desert pavement or underlain by calcrete (i.e., hardpan; e.g., Ritter et al. 1995; French and Miller 2012).
(a) The elevation map and the locations of the Las Vegas Wash (LVW) watershed, rain gauges, and the KESX radar and its 200-km boundary. The land use for LVW watershed in (b) 1950 and (c) 2019. Spatial distribution of river channels (d) before and (e) after the construction of flood conveyance systems and detention basins. Elevation contours in (b) and (c) are shown in 500-m intervals. U.S. Geological Survey land use (USGS-LU) data were used for land use in 1950, whereas the National Land Cover Database (NLCD) was used for land use in 2019.
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
The population in the Las Vegas metropolitan area has increased from less than 35 000 in 1950 to 2.6 million in 2020 (U.S. Census Bureau 2020), leading to substantial urbanization within the LVW watershed. According to land use data (see section 3a for details), developed land has increased from 134 km2 (3.3%) to 998 km2 (24.4%) over the past seven decades, whereas shrubland and barren land have decreased by 644 and 196 km2, respectively (Figs. 1b,c). In addition to land use changes, the Clark County Regional Flood Control District (CCRFCD), which was established in 1985, has implemented flood conveyance systems, including storm drains and detention basins, over the developed valley (Figs. 1d,e). Land use changes and the densely built flood conveyance systems have increased the hydrologic efficiency in the valley, resulting in higher flood peaks. Although these detention basins have successfully stopped flood flows from the mountainous region, they are less effective in mitigating the overall impacts of urbanization at a watershed scale [Fig. 1e; see Yu et al. (2023) for the nonstationary hydrology for the LVW watershed].
3. Data and methods
a. Land use and land cover data
The 1950 and 2019 land use data for the LVW watershed are from the U.S. Geological Survey (USGS; Sohl et al. 2016) and the National Land Cover Database (NLCD; Homer et al. 2020), respectively. The USGS land use (USGS-LU), which is at annual and 250-m resolution over the 1938–92 period, is a model-based land use and land cover “backcast” based on numerous independent historical data sources (Sohl et al. 2016). The NLCD is a remote sensing-based dataset available from 2001 to 2019 at 2–3-yr intervals. Land use from the two datasets was categorized into five classes: developed land, barren land, shrubland, forest, and open water/wetland (Figs. 1b,c).
b. Streamflow data
Based on the instantaneous (i.e., 15-min resolution) streamflow for the USGS gauge at the watershed outlet (USGS: 09419753), we selected the largest 52 flood events during the 1996–2021 period, which averaged two events per year. The selected flood events included all annual peak flows for the period except for 1996, for which the annual peak flow was too small (13 m3 s−1) to be included. Among the largest 52 flood events, 24 occurred during the cool and 28 during the warm seasons, respectively. Of the 28 warm-season events, 25 occurred during the NAM season.
Using instantaneous streamflow, we defined the flooding rise time as the difference between the start and peak times. The start time is when the flow exceeds 15 m3 s−1, which is slightly higher than the dry-season flow of ∼10 m3 s−1 (the result of effluent from two wastewater treatment plants). The flooding rise time has been widely used as a proxy to indicate the “flashiness” of a flood event (e.g., Georgakakos 1986; Gourley et al. 2013; Saharia et al. 2017).
c. Hydrometeorological variables
The meteorological variables used in this study include CG lightning flashes from the National Lightning Detection Network (NLDN; Cummins et al. 1998), rain-gauge-based rainfall data from the CCRFCD, and radar reflectivity fields from the KESX WSR-88D radar (Fig. 1a). Here, we used a hybrid scan that was constructed from the four lowest antenna elevation angles (0.5°, 1.5°, 2.4°, and 3.4°). The LVW watershed was largely unaffected by radar blockage due to the high radar elevation (1509 m MSL) and long distance (100 km) between the mountains and radar. For example, the KESX radar’s 1.5° beam over the Spring Mountains and Sheep Range has an estimated center height of 4050 and 4258 m MSL, respectively, which is higher than the terrain. However, we acknowledge that this beam angle and the distance between the mountain and radar may miss precipitation at lower elevations (e.g., ∼3000 m MSL). All three datasets were collected for the 1996–2021 period.
The CCRFCD maintains an increasingly dense network of rain gauges within the LVW watershed, which started with 58 gauges in 1996 and reached 153 gauges in 2021. These rain gauges are operated under the Automated Local Evaluation in Real Time (ALERT) system, which reports on an irregular interval. Rainfall observations were aggregated or linearly interpolated into a 15-min resolution. In this study, we used a Python toolkit, Py-ART (Helmus and Collis 2016), to convert 3D polar coordinate volume scan reflectivity fields into 3D Cartesian fields, which were further used for storm tracking.
d. TITAN storm tracking
We used the Thunderstorm Identification, Tracking, Analysis, and Nowcasting (TITAN; Dixon and Wiener 1993) algorithm with KESX 3D radar reflectivity to track and examine the storms that produced the largest 52 flood events during the period of interest for the LVW watershed. TITAN has been widely used for analyzing the structural and evolution properties of flood-producing storms (e.g., Wilson and Mueller 1993; Morel and Senesi 2002; Smith et al. 2016, 2019; Yang et al. 2016, 2019). The main procedure of TITAN is summarized as follows:
-
For each radar scan, TITAN identifies multiple contiguous regions in all grids that have reflectivity larger than a defined threshold (Tz) and the volumes of which exceed a threshold (Tυ). In this study, Tz and Tυ are defined as 45 dBZ and 50 km3, respectively, as recommended by previous studies (e.g., Yang et al. 2016, 2019; Smith et al. 2019).
-
For each identified storm cell, TITAN computes characteristics such as the reflectivity-weighted centroid, area, volume, speed, direction, maximum reflectivity, and echo-top height.
-
TITAN determines the storm path for each storm cell by joining it with its counterpart in the successive radar scan that has the shortest distance and similar characteristics (e.g., shape, area, and volume).
4. Results
a. An overview of the selected events
Although the 3 largest flood events occurred in the warm season, 6 out of the 10 largest flood events occurred in the cool season (Table 1). The median and variability of peak flows between warm- and cool-season events are comparable, which highlights the importance of seasonal flooding in the LVW watershed (Fig. 2a). Different seasonality is associated with different flood-generating mechanisms, the understanding of which is important in flood risk assessment (e.g., Barth et al. 2017; Berghuijs et al. 2019; Yu et al. 2019, 2021, 2022). Although flooding rise time (i.e., flashiness) shows different distributions between the two seasons, it indicates that flood response for some cool-season events can be as quick as warm-season ones (Fig. 2b). The NWS in Las Vegas issued flash flood warnings for 26 out of 28 (93%) of the warm-season events and 15 out of 24 (63%) of the cool-season events (Table 1). The lower percentage of warnings in the cool season is partly because flash flood guidance in arid and urban regions is dominated by rainfall intensity (Sweeney 1992). However, low to moderate rainfall intensities can still cause a rapid flood response and high magnitudes due to urbanization, which concentrates flow.
The largest 52 flood events in the LVW watershed and their corresponding hydrometeorological variables. Peak flows are used to sort events in descending order.
The distribution of (a) peak flows, (b) flooding rise time, (c) time lags between rainfall and flood peaks, (d) rainfall total, and (e) hourly and (f) 15-min rainfall intensity for the selected flood events. The shape of these violin plots represents the probability density function. Total, hourly, and 15-min rainfall intensities are for all rain gauges across the events, representing the spatiotemporal variability of rainfall. NOAA Atlas 14-based hourly and 15-min rainfall intensities for 10-, 100-, and 1000-yr recurrence intervals are shown in (e) and (f).
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
The rainfall durations and accumulated depths responsible for cool-season floods are generally larger than the corresponding values for warm-season events (Figs. 2b,d), which is attributable to the different rainfall-generating mechanisms (as detailed in the following subsections). The lag time, which is the delay between peak rainfall and peak discharge, is comparable between seasons, indicating that cool-season floods can be as “flashy” as warm-season floods (Fig. 2c). Later, we show that the short lag time for cool-season events is the result of relatively high initial channel flows caused by the rain prior to the peak rainfall intensity (see section 4d). Both visual inspection (Figs. 2d–f) and two-sample Kolmogorov–Smirnov tests show that rainfall totals are larger in winter, whereas the 15-min and hourly rainfall intensities are larger in summer (p value < 0.05 in both tests). Based on the estimates from the NOAA Atlas 14 (Bonnin et al. 2006), the recurrence intervals for rain-gauge-based hourly and 15-min rainfall intensities can be as large as the 1000-yr recurrence interval in the warm season (Figs. 2e,f). In the cool season, 15-min rainfall intensities never exceed the 100-yr recurrence interval, whereas hourly rainfall intensities can be as large as the 1000-yr recurrence interval (Figs. 2e,f).
The spatial distributions of the maximum 15-min rainfall intensities for the selected cool- and warm-season events are shown in Figs. 3a and 3b, respectively. The maximum 15-min rainfall intensities are typically less than 50 mm h−1 for cool-season events and above 50 mm h−1 for warm-season events (Figs. 3a,b). Notably, the maximum 15-min rainfall intensities in the cool (warm) season show no relationship (a negative relationship) between rainfall intensity and elevation (inset figures in Figs. 3a,b). In sharp contrast, the mean seasonal rainfall climatology shows a marked orographic effect, as shown by two positive linear regressions (p value < 0.05) between mean seasonal rainfall and elevation (Figs. 3c,d).
Maximum 15-min rainfall intensities for each rain gauge in the (a) cool and (b) warm seasons during the 1996–2021 period. Mean (c) cool-season and (d) warm-season total rainfall during the same period. Embedded figures in each panel show the relationship between elevation and both rainfall totals and intensities.
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
In both the cool and warm seasons, selected peak flows have a higher correlation with short-duration rainfall intensities than event rainfall totals, highlighting the role of intense storm cores in driving flood peaks (Fig. 4). The correlations between peak flows and both rainfall totals and intensities range from 0.47 to 0.72 and from 0.70 to 0.85 for cool- and warm-season events, respectively (Fig. 4). In the cool season, both rainfall and floods are modestly correlated with IVT, CAPE, and PW (ρ ≈ 0.3) but correlate least with CG flashes (ρ ≈ 0; Fig. 4a). This is consistent with the characteristics of nonconvective precipitation in the cool season. By comparison, warm-season rainfall and floods have high correlation with CAPE (ρ > 0.7), PW (ρ ≈ 0.45), and CG flashes (ρ ≈ 0.25), which underscores the role of moist and unstable air in driving such events. Notably, warm-season rainfall and floods negatively correlate with IVT because organized thunderstorms with relatively high IVT result in lower rainfall intensity over the LVW watershed compared to scattered, orographic-induced thunderstorms (see sections 4c and 4d for details).
Pearson correlations between peak flows and the corresponding hydroclimatological and meteorological variables for the (a) cool and (b) warm seasons. Hydroclimatological variables include event mean IVT (kg m−1 s−1), CAPE (J kg−1), and precipitable water (PW; mm). Meteorological variables include CG flashes; rainfall totals [Ptotals (mm)]; and maximum 15-min [P15-min (mm h−1)], 30-min [P30-min (mm h−1)], 1-h [P15-min (mm h−1)], and 2-h [P2-h (mm h−1)] rainfall intensities. All correlation values are significant at the 5% level.
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
b. The synoptic patterns for cool- and warm-season floods
We identified four synoptic patterns for the selected flood events using the geopotential heights at 500 hPa and total column PW. Open wave and cutoff lows occur in the cool season, whereas classic and troughing NAM occur in the warm season (Figs. 5 and 6; Table 1). The open wave synoptic feature, which predominately occurs in December–February, shows longwave troughs situated over coastal California, low PW over the southwestern United States, and strong IVTs along the jet streams (Figs. 5a,c). Storms over the LVW watershed during this pattern are fueled by the strong southwesterly IVTs that bypass the Sierra Nevada (Fig. 5c). Another synoptic pattern of cool-season events was identified as cutoff low systems over Southern California, which frequently occurs in the early cool season months of October and November (Fig. 5b). These cutoff lows are often associated with a cyclonic circulation that tends to transport moisture from the Pacific Southwest and Gulf of California to the LVW watershed (Fig. 5d). Based on the Gershunov et al. (2017) AR detection method, we identified five open wave and three cutoff low pattern AR events (Table 1). The three ARs during the cutoff low conditions were situated perpendicular to the Gulf of California (i.e., southwest–northeast direction) and were deflected by the cyclonic circulations toward the LVW watershed.
Composite synoptic conditions for (a),(c) open wave and (b),(d) cutoff low patterns. (top) Composite precipitable water and 500-hPa geopotential height (m; contours). (bottom) Integrated water vapor transport (IVT; kg m−1 s−1) and wind speed at 925 hPa. The seasonality of such patterns is shown in (a) and (b). The Sierra Nevada and LVW watershed are shown in (c) and (d).
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
Composite synoptic conditions for (a),(c) classic and (b),(d) troughing NAM patterns. (top) Composite precipitable water and 500-hPa geopotential height (m; contours). (bottom) Integrated water vapor transport (IVT; kg m−1 s−1) and wind speed at 925 hPa. The seasonality of such patterns is shown in (a) and (b). The Sierra Nevada and LVW watershed are shown in (c) and (d).
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
For the 28 warm-season floods, we identified two NAM-related synoptic patterns: classic and troughing (Fig. 6). The classic NAM is comparable to the Type I NAM pattern of Maddox et al. (1995), which is characterized as high geopotential height east of the Rockies and substantial PW along the Gulf of California (i.e., gulf surge; Figs. 6a,c). The troughing NAM is similar to the Maddox et al. (1980) Type I western flash flood pattern, during which flash flooding typically occurs along the western side of a longwave ridge (Figs. 6b,d). In both classic and troughing NAM patterns, surface winds and IVTs over the LVW watershed and its surroundings are relatively weak, except for the corridor between the Gulf of California and southern Nevada that transports moisture northward (Figs. 6c,d).
The PW and CAPE for the selected events were compared to the corresponding long-term daily mean values for the 1979–2021 period (Fig. 7). The distributions of daily PW and CAPE are skewed toward low values, indicating that the atmospheric environment for the LVW watershed is generally dry and unfavorable for convective storms. It should be noted, however, that CAPE in this study may be nonrepresentative of real conditions because 1) it was calculated as the mean value from the four NARR grids that encompass the LVW watershed, as opposed to the sounding-based value, and 2) the CAPE was calculated in terms of the surface layer rather than the pressure level that produces the most CAPE (i.e., most unstable CAPE).
The relationships between peak flows and the corresponding precipitable water, CAPE, and IVT during the cool and warm seasons. The color and size of symbols denote peak flows and IVT, respectively, for the 52 events. Numbers on the top of the colored symbols denote the rank (i.e., ID) of the selected 52 events. Daily mean precipitable water and CAPE for the 1979–2021 period are shown as gray dots in the scatterplot and their histograms are also shown.
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
The PW, CAPE, and IVT for the selected 52 flood events show distinct cool- and warm-season patterns, without distinct clustering in terms of different synoptic patterns within each season (Fig. 7). The warm-season events, including the largest two floods, featured PW and CAPE exceeding the 75th percentiles of the corresponding long-term climatological values. High PW and CAPE indicate a very moist and unstable atmosphere, which is favorable for flood-producing storms in the arid southwestern United States during the NAM season. However, the warm-season events, including a few of the larger ones (e.g., events 1, 2, 3, 12, and 13), featured relatively small IVT, consistent with the negative correlation between the warm-season floods and IVT (Fig. 4b). This implies that a substantial portion of moisture for warm-season storms has been transported over the watershed prior to rather than during the event. For example, the basin-averaged PW for the record flood on 8 July 1999 increased from 13 mm on 5 July to 34 mm on the event date, with a mean daily increase of 7 mm.
The cool-season events occurred in a narrow range of PW (10–20 mm), but in broad ranges of CAPE (1–400 J kg−1) and IVT (50–500 kg m−1 s−1). The PW for cool-season events was generally low (<20 mm) because of the low water-holding capacity of the cold atmosphere. CAPE was relatively high for the cutoff low events but low for the open wave events. For example, the largest two cool-season floods (events 4 and 5) occurred when the CAPE was relatively high for the season (>200 J kg−1; Fig. 7). Although CAPE in the cool season is too weak to develop deep convection and organized thunderstorms, it is conducive to producing scattered convective cells embedded in frontal systems. Eight out of 24 cool-season events were associated with inland-penetrating ARs (events 4, 5, 7, 9, 10, 14, 16, and 18 in Fig. 7) and typically yielded higher peak flows, which highlights the role of concentrated moisture transport and favorable synoptic dynamics in increasing the efficiency of storm production.
For both cool- and warm-season flood events, there is large variability in the peak flows for given values of CAPE, PW, and IVT (Fig. 7). There are two main reasons for this variability:
-
CAPE, PW, and IVT indicate the potential for flash-flood-producing storms, but they are less predictive of the initiation and intensity of such storms. Other factors such as surface fronts, low-level convergence, convective inhibition (CIN), and vertical wind shear also play important roles in storm strength.
-
The rainfall spatiotemporal structure—including storm location and movement relative to the watershed—can modulate the magnitudes of peak flows at the watershed outlet. For urbanized watersheds, the overlap between storm centers and the impervious developed land, as well as the storm direction relative to the configuration of flood conveyance systems, also determines peak flows (e.g., Smith et al. 2005a,b; Wright et al. 2012).
c. Storm-tracking spatial characteristics
The Lagrangian properties of flash-flood-producing storms were analyzed using the TITAN storm-tracking algorithm combined with KESX 3D radar reflectivity (Fig. 8). In general, storm tracks are concentrated over the low valley (<1000 m MSL), and their directions, velocity, area, duration, and travel distance are linked to their corresponding seasons (Fig. S2 in the online supplemental material).
The storm tracks and the corresponding speed and directions for (a) open wave, (b) cutoff low, (c) NAM isolated thunderstorm, and (d) NAM organized thunderstorm events. The color of the storm tracks denotes the maximum reflectivity of storms.
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
Cool-season storms under the open wave condition are southerly and southwesterly dominated, with relatively high translational velocity and low maximum reflectivity (Fig. 8a). Such storms are driven by strong and moist jet streams transporting moisture from the North Pacific and Gulf of California (Fig. 5c). Storm tracks under cutoff low conditions show south–north and southwest–northeast orientation, driven by cyclonic circulations (Figs. 5d and 8b). They are also associated with relatively high maximum reflectivity due to atmospheric instability. Furthermore, the cool-season events with a long duration typically consist of multiple rain cells passing over the LVW watershed (i.e., the training effect; e.g., Doswell et al. 1996) from the low valley in south to the high terrain in north. Such upslope flows advected by the large-scale frontal system cause more accumulated precipitation depths over the higher elevation [Fig. 3c; Fig. 9a in Yu et al. (2023) showed a similar spatial distribution of winter precipitation based on grided precipitation data]. Therefore, the main driver for cool-season rainfall is the interaction between abundant moisture transport and complex terrain.
In the warm season, classic and troughing NAM patterns are associated with isolated and organized thunderstorms, respectively. The storm tracks of isolated thunderstorms show low translational velocities, short distances, and no dominant direction (Fig. 8c; 8 July 1999 storm in Figs. S3a–d), but higher velocities, longer distances, and southerly and southwesterly directions for organized thunderstorms (Fig. 8d). Isolated, orographic-driven thunderstorms are common features in complex terrain where the radiational heating and cooling of land acts to trigger vertical motion and instability (see McCollum et al. 1995; Javier et al. 2007; Tarolli et al. 2013). The orographic-driven storms typically remain over the mountains until convective outflows push them downslope, which causes heavy rainfall over the low valley (see section 5a for a discussion of the different orographic effects in the LVW watershed). Unlike orographic thunderstorms, organized thunderstorms belong to mesoscale convective systems that affect wide areas over the NAM region (Fig. 7c). For instance, the 26 July 2021 storm was initiated >150 km northeast of the LVW watershed and then moved southwestward, passing across the LVW watershed in four hours (Figs. S3e–h).
Although the LVW watershed has a northwest–southeast orientation (113°; Fig. S3a), its five subwatersheds that drain a large portion of the urban area are southwest–northeast oriented (Figs. S3b–f). Therefore, southwesterly storm tracks in two seasons are consistent with the prevailing directions of highly urbanized subwatersheds, resulting in “storms chasing floods.” Glancy and Harmsen (1975) demonstrated that one important factor that enhanced the severity of the 1974 Eldorado Canyon flash flood in southern Nevada was the eastward downstream motion of thunderstorm cells along the canyon. Furthermore, floods from these subwatersheds tend to arrive in the main wash at approximately the same time because of their comparable size, which creates a “runoff traffic jam.”
Cool-season storms are deemed to be nonconvective given the absence of CG lightning flashes (Table 1; Figs. 9a,b) and low echo-top height (Figs. 10g,h). In contrast, warm-season storms are typically thunderstorms with a large number of CG lightning flashes (Figs. 9c,d) and high echo-top height (Figs. 11g,h). However, some events under the cutoff low pattern showed the existence of a few CG flashes (Table 1 and Fig. 9b). All of those cases occurred during October, which is a transition month between the warm-season convective NAM storms and the cool-season synoptic pattern storms.
Daily mean CG lightning flash density for events that occurred under (a) open wave, (b) cutoff low, (c) NAM isolated thunderstorm, and (d) NAM organized thunderstorm conditions. Black lines denote elevation contour lines with 500-m intervals.
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
Composite time series of cool-season events during the (left) open wave and (right) cutoff low conditions. The time series are in terms of (a),(b) 15-min rainfall rates, (c),(d) peak flows at the watershed outlet, (e),(f) maximum storm reflectivity, and (g),(h) maximum storm echo-top height. Time series are centered on 15-min rainfall peaks. The mean rainfall proportions for the 4-h peak rainfall window, as well as the 10-h periods before and after peak rainfall are shown in (a) and (b). Solid lines in (c) and (d) are the locally weighted scatterplot smoothing (LOWESS) regressions. The largest flood peak discharge values are also shown in (c) and (d). The echo-top height is relative to the elevation of KESX radar, which is 1509 m MSL.
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
Composite time series of warm-season events during the (left) classic and (right) troughing NAM patterns. The time series are in terms of (a),(b) 15-min rainfall rates, (c),(d) peak flows at the watershed outlet, (e),(f) maximum storm reflectivity, and (g),(h) maximum storm echo-top height. Time series are centered on 15-min rainfall peaks. The mean rainfall proportions for the 4-h peak rainfall window, as well as 10-h periods before and after peak rainfall are shown in (a) and (b). Solid lines in (c) and (d) are the locally weighted scatterplot smoothing (LOWESS) regressions. The largest flood peak discharge values are also shown in (c) and (d). The echo-top height is relative to the elevation of KESX radar, which is 1509 m MSL.
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
The spatial distribution of CG lightning flash density in the LVW watershed in the warm season is weakly dependent on terrain, with more lightning occurring over the low valley (Figs. 9c,d). For example, during isolated thunderstorm events, mean flashes exceed 150 over the valley, whereas the mean is less than 20 over the mountains, demonstrating intense convection over the urban area (Fig. 9c). This spatial distribution of lightning flashes for warm-season events is independent of terrain, in contrast to the terrain-dependent pattern of warm-season lightning climatology (Fig. S4). Consistent with the spatial distribution of maximum 15-min rainfall intensities in the warm season (Fig. 3b), high lightning flash density over the valley suggests that a prerequisite for urban flooding is the occurrence of moderate to heavy rainfall over the developed low valley.
d. The evolution of cool- and warm-season events
The evolution of urban flash flooding was examined through a time series of 15-min rainfall intensities, peak flows at the watershed outlet, maximum reflectivity, and echo-top height (Figs. 10 and 11). The two latter features provide insights into the convective intensity over the life cycle of the storm elements. The evolution of these flood events shows three common features: 1) rainfall rates covary with maximum reflectivity and echo-top height of tracked storms, 2) the convective intensity of tracked storms rapidly decays after the rainfall peaks, and 3) there is a relationship between deeper storms and higher-intensity rainfall.
Cool-season storms show long durations, multiple peaks, and low to moderate rainfall intensities (Figs. 10a,b). On average, rainfall peaks within a 4-h window account for 48% and 42% of total rainfall for open wave and cutoff low events, respectively (Figs. 10a,b). Additionally, peak rainfall intensities are slightly higher than the intensities before and after the peaks. Continuous and moderate rainfall intensities result from multiple storms passing through the LVW watershed (i.e., a training effect). The interaction of rainfall with the impervious valley floor results in multiple flood peaks at the watershed outlet, with the largest peak coming last.
Convective intensity as indicated by maximum reflectivity and echo-top height is low for open wave events but moderate for cutoff low events (Figs. 10e–h). The relatively high maximum reflectivity (e.g., >60 dBZ) for some storms during the cutoff low condition can be attributed to the existence of relatively high CAPE. Echo-top height for cool-season storms, especially under the open wave condition, rarely exceeds 5 km above the radar height, and the corresponding maximum storm reflectivity rarely exceeds 55 dBZ. The KESX radar reflectivity associated with peak rainfall rates for one open wave event (Figs. 12a–c) and one cutoff low (Figs. 12d–f) event shows that both storms are very shallow. The differential reflectivity (ZDR), which is the logarithmic ratio of the horizontally to vertically polarized reflectivity, has been widely used to detect updrafts (e.g., Yang et al. 2016, 2019; Smith et al. 2019). ZDR for these two cool-season events indicates the absence of updrafts at the low level (Figs. 12c,f).
(left) The KESX radar reflectivity, (center) its vertical profile for the selected latitudinal or longitudinal slices, and (right) differential reflectivity at (a)–(c) 1511 UTC 9 Jan 2018, (d)–(f) 0247 UTC 13 Mar 2020, (g)–(i) 2146 UTC 30 Jun 2016, and (j)–(l) 0359 UTC 26 Jul 2021. Four events occurred under the open wave, cutoff low, and classic and troughing NAM synoptic patterns (shown from top to bottom, respectively).
Citation: Journal of Hydrometeorology 24, 11; 10.1175/JHM-D-23-0002.1
The flood responses for cool-season events also show multiple peaks, with the streamflow prior to the rainfall peaks being relatively high, leading to large flood volume (Figs. 10c,d). For some cool-season events, the streamflow prior to the rainfall peaks can be as high as 100 m3 s−1, 10 times larger than the dry-period discharge. Rainfall with moderate and low intensity can yield large streamflows prior to the rainfall peaks because of the impervious land and constructed flood conveyance system over the southeastern portion of the LVW watershed (Figs. 1c,e). Therefore, the cool-season peak flows depend on rainfall intensities and volumes, as well as the antecedent channel flows in the flood conveyance system.
On the contrary, warm-season storms mainly occur in the 4-h peak rainfall window and typically have a single rainfall peak (Figs. 11a,b). Rainfall during the 10-h period prior to the rainfall peak has low intensity and contributes a mean of 12% and 16% to rainfall totals for isolated and organized thunderstorms, respectively (Figs. 11a,b). The isolated thunderstorms under the classic NAM pattern show an “explosive” feature with short duration but the most extreme convective intensity (Figs. 11e,g). The organized thunderstorms have a slightly longer life cycle, but a weaker convective intensity compared to the isolated thunderstorms (Figs. 11f,h).
The warm-season storms also reflect high echo-top heights, indicating deep convection. Radar reflectivities associated with rainfall peaks for two warm-season storms show deep and intense convection, and the corresponding ZDR indicates strong updraft at the low level (Figs. 12g–l). It is common in the NAM season that outflows of existing storm cells interact with ambient convective environments, resulting in a strong updraft and new storm cells (e.g., McCollum et al. 1995; Corfidi 2003). For example, the triggering mechanism for the July 1999 storm and its resultant record flood was the interaction between outflows from nighttime storms and the eastern flanks of the Spring Mountains (Li et al. 2003).
The “explosive” flood responses for the warm-season events show a single peak, a time to peak of approximately 4 h, a steep rising limb, and a small flood volume (Figs. 11c,d). Rainfall intensities prior to the peak rainfall rates were too small to yield apparent changes in the streamflow at the watershed outlet (Figs. 11c,d). Another reason for the rapid flood response in the warm season is that the intense storms often occurred over the southeastern portion of the LVW watershed, which is shown by the spatial distribution of storm tracks (Figs. 8c,d) and CG lightning flash density (Figs. 9c,d). The southeastern part of the LVW watershed is highly urbanized and close to the watershed outlet and therefore has rapid runoff conveyance.
5. Discussion
a. Orographic effects under different synoptic conditions
The developed Las Vegas valley is encompassed by the Spring Mountains to the west, the Sheep Range to the north, and a few small mountains to the south and east, which play important roles in influencing the precipitation mechanisms among different seasons (Fig. 1a). During the cool season, under either the open wave or cutoff low synoptic patterns, nonconvective storms are southerly dominant, widespread, and fast-moving and initiate before they cross the LVW watershed (Figs. 8a,b). Such storms readily pass over the small mountains in the south of the LVW watershed and are slightly enhanced on the upwind side via the seeder–feeder mechanism (Bergeron 1965; Passarelli and Boehme 1983). When preexisting precipitating clouds are advected over small mountains at a high elevation, the precipitation particles from the upper cloud (i.e., seeder) are aggregated by the cloud at the lower level (i.e., feeder). These storms will pass over the low valley and approach the higher-elevation Spring Mountains and Sheep Range, where they are forced to flow upward or are partially blocked by the terrain, depending on the strength of steering winds, atmospheric stability, and terrain height. Both terrain blocking and upslope forcing enhance the precipitation accumulation over the upwind side of the mountain (Fig. 3c).
During the warm season, the radiational heating and cooling of the terrain play a critical role in driving the isolated thunderstorms (Fig. 8c). The solar heating of the elevated terrain during the daytime forces upslope valley–mountain circulations, which can sometimes cause air parcels to rise above the level of free convection. Convective storms over the mountains may propagate away from the mountains and their outflows tend to trigger new convective cells within the low valley, which are the principal agents for urban flash flooding at the LVW watershed. On the other hand, nighttime cooling over high terrain causes downslope flow, which may interact with unstable low-level moist air associated with NAM to trigger deep convection within the low valley. For organized thunderstorms, the complex terrain plays a similar role (i.e., seeder–feeder), except that some convective cells may form over the upwind side.
The results here suggest that these orographic effects cannot be described as a linear relation between terrain height and long-term mean climatological precipitation, as is commonly done in many gauge-based precipitation datasets such as Parameter-Elevation Regressions on Independent Slopes Model (PRISM; Daly et al. 1997). Instead, the microphysics, dynamics, and thermodynamics of the moist airflow and their interactions with the geometry of complex terrain tend to affect when, where, and how precipitation falls. For comprehensive reviews of these orographic effects, see Barros and Lettenmaier (1994), Houze (2012), and Smith (2019). Recent advancement and research efforts in radar-based precipitation estimates (e.g., NRC 2005), uncertainty quantification of precipitation products (e.g., Bytheway et al. 2020, 2022), and high-resolution (e.g., convection-permitting) atmospheric models (e.g., Lundquist et al. 2019; Jing et al. 2017) can improve monitoring and forecasting of precipitation over the complex terrain.
b. Recommendations for flash flood management in the Southwest
First, we recommend monitoring of synoptic features and storm movement, which can help increase the accuracy and lead time of flash flood forecasts, watches, and warnings. The observed synoptic features should be used as boundary conditions in high-resolution mesoscale numerical models to improve their forecasting skills. In addition, we recommend leveraging a network of regional WSR-88D radars and remote sensing data, such as Geostationary Operational Environmental Satellite cloud-top temperature, to extend radar coverage and to better understand the movement of convective storms and its interaction with urban area in advance.
Second, we recommend establishing a process-based framework for verifying flash flood warnings that accounts for the various drivers that contribute to flash flooding. Currently, the NWS verifies flash flood warnings by comparing them to storm data reports (“ground truth”) and evaluating the probability of detection and false alarm rates (NRC 2005). However, it is essential to ensure that the forecasting system can accurately detect and simulate each synoptic scale (e.g., cutoff lows), mesoscale (e.g., fronts) as well as hydrological processes involved in flash flooding.
Finally, the warming impacts on both cool-season inland-penetrating ARs and warm-season NAM-associated convective rainfall over the southwestern United States deserve further exploration. Although a warmer future will lead to a more favorable environment for inland-penetrating ARs, the landfalling locations of future ARs are less certain (Yin 2005; Chang et al. 2012; Payne et al. 2020). Similarly, climate warming can enhance Gulf surges during the NAM season whereas a projected decrease in the relative humidity will cause a more stable atmosphere, thus inhibiting deep convection (Pascale et al. 2019; Fowler et al. 2021). The increasing availability of higher resolution model experiments (e.g., Prein et al. 2017) and large ensembles (e.g., Ma et al. 2020) will facilitate these explorations.
6. Summary and conclusions
We derived a catalog of the flash-flood-producing storms that caused the 52 largest flood peaks over the 1996–2021 period for the LVW watershed outlet. We investigated these events considering their synoptic-scale weather patterns, storm-scale structures, and evolutions using NARR reanalysis, gauge-based rainfall observations, 3D radar reflectivity fields, CG lightning flash counts, and USGS streamflow observations. The main conclusions are summarized below.
-
The flash-flood-producing storms for the LVW watershed occurred in two pronounced seasons, both of which result in comparable peak flows at the watershed outlet (Fig. 2). Cool-season storms are thermodynamically driven with long-duration and low- to moderate-intensity rainfall, whereas warm-season storms are dynamically driven with short-duration and high-intensity rainfall (Figs. 4 and 7). Cool-season storms are associated with open wave and cutoff low synoptic patterns that determine the direction and intensity of IVT over the Pacific Southwest. Such IVT, especially when exceeding the AR threshold (250 kg m−1 s−1), plays a key role in determining the rainfall intensity and flood magnitude in the cool season. Warm-season storms are associated with two distinctive upper-level monsoonal ridges and are highly correlated with the CAPE and PW.
-
We used the TITAN storm-tracking algorithm and KESX radar reflectivity to examine the Lagrangian properties of storms at a small scale. Storm motion is closely tied to synoptic patterns (Fig. 8). In the cool season, storms under the open wave pattern are southerly and southwesterly dominated, whereas storms under the cutoff low pattern are southeasterly dominated. In the warm season, storms under the classic pattern show no apparent direction, whereas storms under the troughing NAM pattern are southerly and southwesterly dominated. High values of maximum reflectivity in both seasons are concentrated over the developed low valley, consistent with the spatial distribution of maximum 15-min rainfall rates. The occurrence of heavy rainfall over the low valley is an important prerequisite for high peak flows at the watershed outlet, highlighting the importance of urbanization on flash flood runoff production.
-
Cool- and warm-season storms show contrasting convective properties in terms of CG lightning flashes, maximum reflectivity, updrafts, and echo-top height. Cool-season storms are mostly nonconvective with an absence of CG flashes and updrafts, moderate maximum reflectivity, and low echo-top height with the exception of a few convective storms that do occur under cutoff low conditions in the October “shoulder” (i.e., seasonal transition) period. On the contrary, summer storms are convective with substantial CG lightning flashes and high maximum reflectivity and echo-top heights. In particular, localized orographic-induced thunderstorms with deep convection over the low valley drive the upper-tail floods for the LVW watershed (Figs. 8c and 9c). Furthermore, radiative and dynamic blocking and lifting forcings control the orographic effects on precipitation in cool and warm seasons, respectively.
-
The evolution of flash floods was examined through the aggregated time series of rainfall rates, convective properties of storms, and the hydrographs at the watershed outlet (Figs. 10 and 11). Rainfall rates and the resulting streamflow hydrographs show single peaks in the warm season and multiple peaks in the cool season. The low to moderate rainfall prior to rainfall peaks in the cool season interacts with the impervious and extensively storm-drained urban area, causing relatively high antecedent channel flow that further contributes to peak flows. In the warm season, the isolated thunderstorms reflect an “explosive” feature with sharp increases in rainfall rates, convective intensity, and streamflows (e.g., Smith et al. 2019), whereas organized thunderstorms consist of multiple storms with moderate convective intensity. The evolution of flash flood events during the warm season highlights that a spectrum of storms can result in urban flash floods, although less focus has been given to cool season events. The empirical analyses performed for the LVW watershed can be readily transferred to other urban watersheds in the southwestern United States to better understand the key drivers for urban flash flooding.
Acknowledgments.
G. Y.’s contribution was supported by the Desert Research Institute Maki Postdoctoral Fellowship and Urban Flood Demonstration Program by the Department of the Army (Agreement W912HZ1920011), which also supports J. M., and M. B. B. H.’s and D. B. W’s contributions were supported by the Desert Research Institute and University of Wisconsin–Madison, respectively. We thank Nicole Damon at Desert Research Institute for her assistance with technical editing.
Data availability statement.
USGS-LU and NLCD land use data are available at https://doi.org/10.5066/F7KK99RR and https://www.mrlc.gov/, respectively. NARR reanalysis is available from the NOAA Physical Sciences Laboratory at https://psl.noaa.gov/data/gridded/data.narr.html. Rain gauge and flood conveyance data are available from the Clark County Regional Flood Control District at https://www.regionalflood.org/programs-services/rainfall-and-weather/historical-rainfall/gage-data. KESX radar data are from the NOAA National Centers for Environmental Information at https://www.ncdc.noaa.gov/nexradinv/. Annual peak flow data for all the stream gauges within the Las Vegas Wash watershed are available from USGS at https://waterdata.usgs.gov/nwis.
REFERENCES
Albano, C. M., M. D. Dettinger, and A. A. Harpold, 2020: Patterns and drivers of atmospheric river precipitation and hydrologic impacts across the western United States. J. Hydrometeor., 21, 143–159, https://doi.org/10.1175/JHM-D-19-0119.1.
Ashley, S. T., and W. S. Ashley, 2008: Flood fatalities in the United States. J. Appl. Meteor. Climatol., 47, 805–818, https://doi.org/10.1175/2007JAMC1611.1.
Barlow, M., and Coauthors, 2019: North American extreme precipitation events and related large-scale meteorological patterns: A review of statistical methods, dynamics, modeling, and trends. Climate Dyn., 53, 6835–6875, https://doi.org/10.1007/s00382-019-04958-z.
Barros, A. P., and D. P. Lettenmaier, 1994: Dynamic modeling of orographically induced precipitation. Rev. Geophys., 32, 265–284, https://doi.org/10.1029/94RG00625.
Barth, N. A., G. Villarini, M. A. Nayak, and K. White, 2017: Mixed populations and annual flood frequency estimates in the western United States: The role of atmospheric rivers. Water Resour. Res., 53, 257–269, https://doi.org/10.1002/2016WR019064.
Bergeron, T., 1965: On the low-level redistribution of atmospheric water caused by orography. Proc. Int. Conf. on Cloud Physics, Tokyo, Japan, International Association of Meteorology and Atmospheric Physics, 96–100.
Berghuijs, W. R., S. Harrigan, P. Molnar, L. J. Slater, and J. W. Kirchner, 2019: The relative importance of different flood-generating mechanisms across Europe. Water Resour. Res., 55, 4582–4593, https://doi.org/10.1029/2019WR024841.
Bonnin, G. M., D. Martin, B. Lin, T. Parzybok, M. Yekta, and D. Riley, 2006: Semiarid Southwest (Arizona, Southeast California, Nevada, New Mexico, Utah). Vol. 1, Precipitation-Frequency Atlas of the United States, NOAA Atlas 14, Vol. 1, Version 5.0, 265 pp., https://www.weather.gov/media/owp/hdsc_documents/Atlas14_Volume1.pdf.
Bytheway, J. L., M. Hughes, K. Mahoney, and R. Cifelli, 2020: On the uncertainty of high-resolution hourly quantitative precipitation estimates in California. J. Hydrometeor., 21, 865–879, https://doi.org/10.1175/JHM-D-19-0160.1.
Bytheway, J. L., M. Hughes, R. Cifelli, K. Mahoney, and J. M. English, 2022: Demonstrating a probabilistic quantitative precipitation estimate for evaluating precipitation forecasts in complex terrain. Wea. Forecasting, 37, 45–64, https://doi.org/10.1175/WAF-D-21-0074.1.
Chang, E. K. M., Y. Guo, and X. Xia, 2012: CMIP5 multimodel ensemble projection of storm track change under global warming. J. Geophys. Res., 117, D23118, https://doi.org/10.1029/2012JD018578.
Collier, C. G., 2007: Flash flood forecasting: What are the limits of predictability? Quart. J. Roy. Meteor. Soc., 133, 3–23, https://doi.org/10.1002/qj.29.
Corfidi, S. F., 2003: Cold pools and MCS propagation: Forecasting the motion of downwind-developing MCSs. Wea. Forecasting, 18, 997–1017, https://doi.org/10.1175/1520-0434(2003)018<0997:CPAMPF>2.0.CO;2.
Cummins, K. L., E. P. Krider, and M. D. Malone, 1998: The US National Lightning Detection Network/sup TM/and applications of cloud-to-ground lightning data by electric power utilities. IEEE Trans. Electromagn. Compat., 40, 465–480, https://doi.org/10.1109/15.736207.
Daly, C., G. H. Taylor, and W. P. Gibson, 1997: The PRISM approach to mapping precipitation and temperature. Proc. 10th AMS Conf. on Applied Climatology, Reno, NV, Amer. Meteor. Soc., 20–23, https://prism.oregonstate.edu/documents/pubs/1997appclim_PRISMapproach_daly.pdf.
Davis, R. S., 2001: Flash flood forecast and detection methods. Severe Convective Storms, Meteor. Monogr., No. 28, Amer. Meteor. Soc., 481–526.
Dettinger, M. D., F. M. Ralph, T. Das, P. J. Neiman, and D. R. Cayan, 2011: Atmospheric rivers, floods and the water resources of California. Water, 3, 445–478, https://doi.org/10.3390/w3020445.
Dixon, M., and G. Wiener, 1993: TITAN: Thunderstorm Identification, Tracking, Analysis, and Nowcasting—A radar-based methodology. J. Atmos. Oceanic Technol., 10, 785–797, https://doi.org/10.1175/1520-0426(1993)010<0785:TTITAA>2.0.CO;2.
Doswell, C. A., III, H. E. Brooks, and R. A. Maddox, 1996: Flash flood forecasting: An ingredients-based methodology. Wea. Forecasting, 11, 560–581, https://doi.org/10.1175/1520-0434(1996)011<0560:FFFAIB>2.0.CO;2.
Fowler, H. J., C. Wasko, and A. F. Prein, 2021: Intensification of short-duration rainfall extremes and implications for flood risk: Current state of the art and future directions. Philos. Trans. Roy. Soc., A379, 20190541, https://doi.org/10.1098/rsta.2019.0541.
French, R. H., and J. J. Miller, 2012: Flood Hazard Identification and Mitigation in Semi- and Arid Environments. World Scientific, 223 pp.
Fulton, R. A., J. P. Breidenbach, D.-J. Seo, D. A. Miller, and T. O’Bannon, 1998: The WSR-88D rainfall algorithm. Wea. Forecasting, 13, 377–395, https://doi.org/10.1175/1520-0434(1998)013<0377:TWRA>2.0.CO;2.
Georgakakos, K. P., 1986: On the design of national, real-time warning systems with capability for site-specific, flash-flood forecasts. Bull. Amer. Meteor. Soc., 67, 1233–1239, https://doi.org/10.1175/1520-0477(1986)067<1233:OTDONR>2.0.CO;2.
Gershunov, A., T. Shulgina, F. M. Ralph, D. A. Lavers, and J. J. Rutz, 2017: Assessing the climate-scale variability of atmospheric rivers affecting western North America. Geophys. Res. Lett., 44, 7900–7908, https://doi.org/10.1002/2017GL074175.
Glancy, P. A., and L. Harmsen, 1975: A hydrologic assessment of the September 14, 1974, flood in Eldorado Canyon, Nevada. USGS Professional Paper 930, 28 pp., https://pubs.usgs.gov/publication/pp930.
Gourley, J. J., and Coauthors, 2013: A unified flash flood database across the United States. Bull. Amer. Meteor. Soc., 94, 799–805, https://doi.org/10.1175/BAMS-D-12-00198.1.
Hapuarachchi, H. A. P., Q. J. Wang, and T. C. Pagano, 2011: A review of advances in flash flood forecasting. Hydrol. Processes, 25, 2771–2784, https://doi.org/10.1002/hyp.8040.
Hatchett, B. J., S. Burak, J. J. Rutz, N. S. Oakley, E. H. Bair, and M. L. Kaplan, 2017: Avalanche fatalities during atmospheric river events in the western United States. J. Hydrometeor., 18, 1359–1374, https://doi.org/10.1175/JHM-D-16-0219.1.
Hatchett, B. J., and Coauthors, 2020: Observations of an extreme atmospheric river storm with a diverse sensor network. Earth Space Sci., 7, e2020EA001129, https://doi.org/10.1029/2020EA001129.
Heiss, W. H., D. L. McGrew, and D. Sirmans, 1990: NEXRAD: Next generation weather radar (WSR-88D). Microwave J., 33, 79–89.
Helmus, J. J., and S. M. Collis, 2016: The Python ARM radar toolkit (Py-ART), a library for working with weather radar data in the Python programming language. J. Open Res. Software, 4, e25, https://doi.org/10.5334/jors.119.
Hodgkins, G. A., R. W. Dudley, S. A. Archfield, and B. Renard, 2019: Effects of climate, regulation, and urbanization on historical flood trends in the United States. J. Hydrol., 573, 697–709, https://doi.org/10.1016/j.jhydrol.2019.03.102.
Homer, C., and Coauthors, 2020: Conterminous United States land cover change patterns 2001–2016 from the 2016 national land cover database. ISPRS J. Photogramm. Remote Sens., 162, 184–199, https://doi.org/10.1016/j.isprsjprs.2020.02.019.
Houze, R. A., Jr., 2012: Orographic effects on precipitating clouds. Rev. Geophys., 50, RG1001, https://doi.org/10.1029/2011RG000365.
Javier, J. R. N., J. A. Smith, J. England, M. L. Baeck, M. Steiner, and A. A. Ntelekos, 2007: Climatology of extreme rainfall and flooding from orographic thunderstorm systems in the upper Arkansas River basin. Water Resour. Res., 43, W10410, https://doi.org/10.1029/2006WR005093.
Jing, X., B. Geerts, Y. Wang, and C. Liu, 2017: Evaluating seasonal orographic precipitation in the interior western United States using gauge data, gridded precipitation estimates, and a regional climate simulation. J. Hydrometeor., 18, 2541–2558, https://doi.org/10.1175/JHM-D-17-0056.1.
Leopold, L. B., 1968: Hydrology for urban land planning—A guidebook on the hydrologic effects of urban land use. USGS Circular 554, 26 pp., https://doi.org/10.3133/cir554.
Li, J., R. A. Maddox, X. Gao, S. Sorooshian, and K. Hsu, 2003: A numerical investigation of storm structure and evolution during the July 1999 Las Vegas flash flood. Mon. Wea. Rev., 131, 2038–2059, https://doi.org/10.1175/1520-0493(2003)131<2038:ANIOSS>2.0.CO;2.
Li, Z., S. Gao, M. Chen, J. J. Gourley, C. Liu, A. F. Prein, and Y. Hong, 2022: The conterminous United States are projected to become more prone to flash floods in a high-end emissions scenario. Commun. Earth Environ., 3, 86, https://doi.org/10.1038/s43247-022-00409-6.
Lucas, R. E., Jr., 2004: Life earnings and rural‐urban migration. J. Political Econ., 112 (Suppl. 1), S29–S59, https://doi.org/10.1086/379942.
Lundquist, J., M. Hughes, E. Gutmann, and S. Kapnick, 2019: Our skill in modeling mountain rain and snow is bypassing the skill of our observational networks. Bull. Amer. Meteor. Soc., 100, 2473–2490, https://doi.org/10.1175/BAMS-D-19-0001.1.
Ma, W., J. Norris, and G. Chen, 2020: Projected changes to extreme precipitation along North American West Coast from the CESM large ensemble. Geophys. Res. Lett., 47, e2019GL086038, https://doi.org/10.1029/2019GL086038.
Maddox, R. A., C. F. Chappell, and L. R. Hoxit, 1979: Synoptic and meso-α scale aspects of flash flood events. Bull. Amer. Meteor. Soc., 60, 115–123, https://doi.org/10.1175/1520-0477-60.2.115.
Maddox, R. A., F. Canova, and L. R. Hoxit, 1980: Meteorological characteristics of flash flood events over the western United States. Mon. Wea. Rev., 108, 1866–1877, https://doi.org/10.1175/1520-0493(1980)108<1866:MCOFFE>2.0.CO;2.
Maddox, R. A., D. M. McCollum, and K. W. Howard, 1995: Large-scale patterns associated with severe summertime thunderstorms over central Arizona. Wea. Forecasting, 10, 763–778, https://doi.org/10.1175/1520-0434(1995)010<0763:LSPAWS>2.0.CO;2.
Mazon, J. J., C. L. Castro, D. K. Adams, H.-I. Chang, C. M. Carrillo, and J. J. Brost, 2016: Objective climatological analysis of extreme weather events in Arizona during the North American monsoon. J. Appl. Meteor. Climatol., 55, 2431–2450, https://doi.org/10.1175/JAMC-D-16-0075.1.
Mazumdar, D., 1987: Rural-urban migration in developing countries. Handbook of Regional and Urban Economics, Vol. 2, Elsevier, 1097–1128, https://doi.org/10.1016/S1574-0080(87)80014-7.
McCollum, D. M., R. A. Maddox, and K. W. Howard, 1995: Case study of a severe mesoscale convective system in Central Arizona. Wea. Forecasting, 10, 643–665, https://doi.org/10.1175/1520-0434(1995)010<0643:CSOASM>2.0.CO;2.
Mesinger, F., and Coauthors, 2006: North American Regional Reanalysis. Bull. Amer. Meteor. Soc., 87, 343–360, https://doi.org/10.1175/BAMS-87-3-343.
Morel, C., and S. Senesi, 2002: A climatology of mesoscale convective systems over Europe using satellite infrared imagery. I: Methodology. Quart. J. Roy. Meteor. Soc., 128, 1953–1971, https://doi.org/10.1256/003590002320603485.
Morin, E., D. C. Goodrich, R. A. Maddox, X. Gao, H. V. Gupta, and S. Sorooshian, 2006: Spatial patterns in thunderstorm rainfall events and their coupling with watershed hydrological response. Adv. Water Resour., 29, 843–860, https://doi.org/10.1016/j.advwatres.2005.07.014.
NCEI, 2022: U.S. billion-dollar weather and climate disasters 1980–present. NOAA/NCEI, https://www.ncei.noaa.gov/access/billions/.
Neiman, P. J., F. M. Ralph, G. A. Wick, J. D. Lundquist, and M. D. Dettinger, 2008: Meteorological characteristics and overland precipitation impacts of atmospheric rivers affecting the west coast of North America based on eight years of SSM/I satellite observations. J. Hydrometeor., 9, 22–47, https://doi.org/10.1175/2007JHM855.1.
NRC, 2005: Flash Flood Forecasting over Complex Terrain: With an Assessment of the Sulphur Mountain NEXRAD in Southern California. National Academies Press, 206 pp.
Oakley, N. S., J. T. Lancaster, M. L. Kaplan, and F. M. Ralph, 2017: Synoptic conditions associated with cool season post-fire debris flows in the transverse ranges of Southern California. Nat. Hazards, 88, 327–354, https://doi.org/10.1007/s11069-017-2867-6.
Ogden, F. L., H. O. Sharif, S. U. S. Senarath, J. A. Smith, M. L. Baeck, and J. R. Richardson, 2000: Hydrologic analysis of the Fort Collins, Colorado, flash flood of 1997. J. Hydrol., 228, 82–100, https://doi.org/10.1016/S0022-1694(00)00146-3.
Pascale, S., L. M. V. Carvalho, D. K. Adams, C. L. Castro, and I. F. A. Cavalcanti, 2019: Current and future variations of the monsoons of the Americas in a warming climate. Curr. Climate Change Rep., 5, 125–144, https://doi.org/10.1007/s40641-019-00135-w.
Passarelli, R. E., Jr., and H. Boehme, 1983: The orographic modulation of pre-warm-front precipitation in southern New England. Mon. Wea. Rev., 111, 1062–1070, https://doi.org/10.1175/1520-0493(1983)111<1062:TOMOPW>2.0.CO;2.
Payne, A. E., and Coauthors, 2020: Responses and impacts of atmospheric rivers to climate change. Nat. Rev. Earth Environ., 1, 143–157, https://doi.org/10.1038/s43017-020-0030-5.
Peleg, N., and E. Morin, 2012: Convective rain cells: Radar-derived spatiotemporal characteristics and synoptic patterns over the eastern Mediterranean. J. Geophys. Res., 117, D15116, https://doi.org/10.1029/2011JD017353.
Prein, A. F., R. M. Rasmussen, K. Ikeda, C. Liu, M. P. Clark, and G. J. Holland, 2017: The future intensification of hourly precipitation extremes. Nat. Climate Change, 7, 48–52, https://doi.org/10.1038/nclimate3168.
Randerson, D., 1976: Meteorological analysis for the Las Vegas, Nevada, flood of 3 July 1975. Mon. Wea. Rev., 104, 719–727, https://doi.org/10.1175/1520-0493(1976)104<0719:MAFTLV>2.0.CO;2.
Ritter, D. F., R. C. Kochel, and J. R. Miller, 1995: Process Geomorphology. 3rd ed. W. C. Brown, 546 pp.
Rivera, E. R., F. Dominguez, and C. L. Castro, 2014: Atmospheric rivers and cool season extreme precipitation events in the Verde River basin of Arizona. J. Hydrometeor., 15, 813–829, https://doi.org/10.1175/JHM-D-12-0189.1.
Rutz, J. J., W. J. Steenburgh, and F. M. Ralph, 2014: Climatological characteristics of atmospheric rivers and their inland penetration over the western United States. Mon. Wea. Rev., 142, 905–921, https://doi.org/10.1175/MWR-D-13-00168.1.
Rutz, J. J., W. J. Steenburgh, and F. M. Ralph, 2015: The inland penetration of atmospheric rivers over western North America: A Lagrangian analysis. Mon. Wea. Rev., 143, 1924–1944, https://doi.org/10.1175/MWR-D-14-00288.1.
Saharia, M., P.-E. Kirstetter, H. Vergara, J. J. Gourley, Y. Hong, and M. Giroud, 2017: Mapping flash flood severity in the United States. J. Hydrometeor., 18, 397–411, https://doi.org/10.1175/JHM-D-16-0082.1.
Smith, B. K., J. Smith, and M. L. Baeck, 2016: Flash flood–producing storm properties in a small urban watershed. J. Hydrometeor., 17, 2631–2647, https://doi.org/10.1175/JHM-D-16-0070.1.
Smith, J. A., A. J. Miller, M. L. Baeck, P. A. Nelson, G. T. Fisher, and K. L. Meierdiercks, 2005a: Extraordinary flood response of a small urban watershed to short-duration convective rainfall. J. Hydrometeor., 6, 599–617, https://doi.org/10.1175/JHM426.1.
Smith, J. A., M. L. Baeck, K. L. Meierdiercks, P. A. Nelson, A. J. Miller, and E. J. Holland, 2005b: Field studies of the storm event hydrologic response in an urbanizing watershed. Water Resour. Res., 41, W10413, https://doi.org/10.1029/2004WR003712.
Smith, J. A., M. L. Baeck, L. Yang, J. Signell, E. Morin, and D. C. Goodrich, 2019: The paroxysmal precipitation of the desert: Flash floods in the southwestern United States. Water Resour. Res., 55, 10 218–10 247, https://doi.org/10.1029/2019WR025480.
Smith, R. B., 2019: 100 years of progress on mountain meteorology research. A Century of Progress in Atmospheric and Related Sciences: Celebrating the American Meteorological Society Centennial, Meteor. Monogr., No. 59, Amer. Meteor. Soc., https://doi.org/10.1175/AMSMONOGRAPHS-D-18-0022.1.
Sohl, T., R. Reker, M. Bouchard, K. Sayler, J. Dornbierer, S. Wika, R. Quenzer, and A. Friesz, 2016: Modeled historical land use and land cover for the conterminous United States. J. Land Use Sci., 11, 476–499, https://doi.org/10.1080/1747423X.2016.1147619.
Sweeney, T. L., 1992: Modernized areal flash flood guidance. NOAA Tech. Memo. NWS HYDRO 44, 37 pp., https://repository.library.noaa.gov/view/noaa/13498.
Tarolli, M., M. Borga, D. Zoccatelli, C. Bernhofer, N. Jatho, and F. al Janabi, 2013: Rainfall space-time organization and orographic control on flash flood response: The Weisseritz event of August 13, 2002. J. Hydrol. Eng., 18, 183–193, https://doi.org/10.1061/(ASCE)HE.1943-5584.0000569.
U.S. Census Bureau, 2020: 2020 population and housing state data. https://www.census.gov/library/visualizations/interactive/2020-population-and-housing-state-data.html.
Wilson, J. W., and C. K. Mueller, 1993: Nowcasts of thunderstorm initiation and evolution. Wea. Forecasting, 8, 113–131, https://doi.org/10.1175/1520-0434(1993)008<0113:NOTIAE>2.0.CO;2.
Wright, D. B., J. A. Smith, G. Villarini, and M. L. Baeck, 2012: Hydroclimatology of flash flooding in Atlanta. Water Resour. Res., 48, W04524, https://doi.org/10.1029/2011WR011371.
Yang, L., J. Smith, M. L. Baeck, B. Smith, F. Tian, and D. Niyogi, 2016: Structure and evolution of flash flood producing storms in a small urban watershed. J. Geophys. Res. Atmos., 121, 3139–3152, https://doi.org/10.1002/2015JD024478.
Yang, L., J. Smith, M. L. Baeck, and E. Morin, 2019: Flash flooding in arid/semiarid regions: Climatological analyses of flood-producing storms in Central Arizona during the North American monsoon. J. Hydrometeor., 20, 1449–1471, https://doi.org/10.1175/JHM-D-19-0016.1.
Yin, J. H., 2005: A consistent poleward shift of the storm tracks in simulations of 21st century climate. Geophys. Res. Lett., 32, L18701, https://doi.org/10.1029/2005GL023684.
Yu, G., D. B. Wright, Z. Zhu, C. Smith, and K. D. Holman, 2019: Process-based flood frequency analysis in an agricultural watershed exhibiting nonstationary flood seasonality. Hydrol. Earth Syst. Sci., 23, 2225–2243, https://doi.org/10.5194/hess-23-2225-2019.
Yu, G., D. B. Wright, and K. D. Holman, 2021: Connecting hydrometeorological processes to low-probability floods in the mountainous Colorado Front Range. Water Resour. Res., 57, e2021WR029768, https://doi.org/10.1029/2021WR029768.
Yu, G., D. B. Wright, and F. V. Davenport, 2022: Diverse physical processes drive upper-tail flood quantiles in the US Mountain West. Geophys. Res. Lett., 49, e2022GL098855, https://doi.org/10.1029/2022GL098855.
Yu, G., J. J. Miller, B. J. Hatchett, M. Berli, D. B. Wright, C. McDougall, and Z. Zhihua, 2023: The nonstationary flood hydrology of an urbanizing arid watershed. J. Hydrometeor., 24, 87–104, https://doi.org/10.1175/JHM-D-22-0117.1.
Zhu, Z., Z. Chen, X. Chen, and G. Yu, 2019: An assessment of the hydrologic effectiveness of low impact development (LID) practices for managing runoff with different objectives. J. Environ. Manage., 231, 504–514, https://doi.org/10.1016/j.jenvman.2018.10.046.