1. Introduction
Florida’s summer climate is characterized by almost daily convective precipitation largely associated with local sea-breeze circulations (Byers and Rodebush 1948). In fact, over parts of peninsular Florida more than 60% of the annual precipitation falls between the months of June and September (Duever et al. 1994). Warm-season rainfall sustains critical groundwater resources relied upon by agricultural and municipal users and supports the diverse flora and fauna endemic to the region (Duever et al. 1994; Obeysekera et al. 1999; Price and Swart 2006). At the same time, this convective precipitation typically occurs with frequent lightning that poses serious risks to people engaged in outdoor work and recreation. Roeder et al. (2015) noted that Florida has the highest lightning fatality rate in the United States, with high lightning flash densities overlapping with regions of high population density. For these reasons, understanding the factors that control the spatial and temporal variability in Florida’s warm-season convective precipitation remains important for those engaged in weather and climate prediction within the state.
While the climatological probability of precipitation rarely falls below 20% anywhere in the state during the months of June, July, and August (JJA), interactions between mesoscale sea-breeze circulations and the synoptic-scale low-level wind result in specific regions of elevated rainfall chances on any given day (Gentry and Moore 1954). For example, synoptic-scale westerly or southwesterly flow in the near-surface layer typically enhances low-level convergence along the east coast sea-breeze front and thus increases the likelihood of precipitation along the eastern Florida peninsula. Arritt (1993) found that the location of enhanced sea-breeze convergence also depends in part on the synoptic-scale wind speed, with slower winds allowing farther-inland progression of the opposing sea-breeze front. Other factors that control warm-season precipitation over Florida include disturbances such as tropical cyclones and tropical easterly waves, frictional convergence associated with the coastline, enhanced or reduced convergence associated with curved coastlines, lake breezes, soil moisture, sea surface temperatures in the Gulf of Mexico and western Atlantic Ocean, and the occasional nontropical synoptic-scale disturbance (e.g., Burpee and Lahiffi 1984; Baker et al. 2001; Misra et al. 2011; Misra and Mishra 2016; Klavans et al. 2020).
National Weather Service (NWS) Weather Forecast Offices (WFOs) in Florida have long recognized the key role of sea-breeze–synoptic flow interactions in controlling daily rainfall patterns over Florida during the warm season. Forecasters at the NWS WFO Tampa Bay developed nine sea-breeze regimes derived entirely from the 1200 UTC 1000–700-hPa layer mean vector wind near Tampa, Florida (Mroczka 2021). These sea-breeze regimes are incorporated into the suite of operational guidance tools used by forecasters to aid in predicting the timing and probability of precipitation within the Tampa Bay NWS WFO county warning area during the summer. Mroczka (2021) noted that during the summer months, the 1200 UTC layer mean 1000–700-hPa wind speed and direction at any point in Florida varies with the position and strength of the subtropical ridge relative to that location. Thus, the western ridge of the North Atlantic subtropical high (NASH) modulates daily rainfall patterns over Florida during the summer months.
Numerous prior studies have investigated the influence of the NASH western ridge on warm-season precipitation throughout the southeastern United States (e.g., Li et al. 2011; Diem 2013; Li et al. 2012; Wei et al. 2019; Nieto Ferreira and Rickenbach 2020; Zorzetto and Li 2021). Li et al. (2012) directly linked interannual variability in summer rainfall over the southeastern United States to variations in the location and intensity of the NASH. Specifically, they found that JJA precipitation over the southeastern United States increased when the position of the NASH western ridge shifted to the south and west of its climatological position, though we note that while this was broadly true for the entire region, the relationship varied along the Florida peninsula. Their results suggested that a southwest shift in the position of the ridge typically coincided with a uniform intensification and expansion of the NASH. They proposed that this expansion and intensification was reinforced through a thermodynamic response in which stronger easterly flow associated with an intensified subtropical high increases surface moisture fluxes via enhanced evaporation. This enhanced evaporation depresses sea surface temperature anomalies and in turn reinforces ridging over the subtropics. Importantly, Li et al. (2011) identified a significant intensification of the NASH and an associated westward shift in the mean position of the western ridge over the period 1948–2007. They attributed this trend to anthropogenic warming, which suggests that anticipated further warming in the coming decades would result in farther-westward migration of the mean JJA NASH western ridge.
While prior research examined the broad relationship between changes in the NASH western ridge and precipitation variability over the southeastern United States, the present study aims to identify whether changes in the NASH have been substantial enough to alter the sea-breeze regime climatology and associated precipitation patterns over Florida. The NWS WFO in Tampa developed the sea-breeze regime methodology for the 18-yr period from 2002 to 2019, but a full 30-yr sea-breeze regime climatology is not available in the published literature or via online resources from the NWS WFO in Tampa. Therefore, an additional purpose of this research is to generate a full 30-yr climatology of the sea-breeze regime distribution to calculate more robust regime statistics and assess the dynamic and thermodynamic environments associated with each regime. While prior research using the sea-breeze regimes primarily focused on short-term weather prediction applications (e.g., Chavez et al. 2022), this study seeks to evaluate sea-breeze regimes on climate time scales. In particular, our primary goals are as follows:
-
Using the methodology employed by the NWS WFO in Tampa, develop a full 30-yr sea-breeze regime climatology and examine the dynamic and thermodynamic environments associated with each regime.
-
Examine interannual variability and long-term trends in the sea-breeze regime distribution to determine whether previously noted changes in the mean position of the NASH western ridge have impacted Florida’s sea-breeze regime climatology.
-
Assess the climatological relationship between interannual variability in sea-breeze regime frequencies and interannual variability in precipitation patterns over Florida to better understand the potential impact of the sea-breeze regime distribution on rainfall occurrence throughout the peninsula on time scales relevant to climate prediction.
2. Data and methods
a. Data
We utilize hourly data from the fifth major global reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ECMWF) (ERA5; Hersbach et al. 2020) for calculation of sea-breeze regimes and analysis of the atmospheric conditions associated with each regime. We first compute a full 30-yr climatology for the period 1991–2020 to summarize the present sea-breeze regime distribution. We compare this present-day regime climatology with the most recent nonoverlapping 30-yr climatological period, 1961–90.
As in Li et al. (2011), we use precipitation data from the Climate Prediction Center (CPC) Unified Gauge-Based Analysis of Daily Precipitation over the CONUS (Xie et al. 2007; Chen et al. 2008). These data provide daily liquid-equivalent precipitation totals (mm day−1) on a 0.25° latitude–longitude grid. The original NWS sea-breeze regime study, Mroczka (2021), made use of the National Centers for Environmental Prediction (NCEP) Stage IV Quantitative Precipitation Estimates dataset, which provides hourly gridded estimates of precipitation at a much higher spatial resolution of 4 km. However, because the Stage IV data are not available before 2002, we traded spatial resolution for a dataset with a longer span to allow for the development of a full 30-yr climatology and assessment of 60-yr trends.
Because we focus on precipitation resulting from sea-breeze interactions with the large-scale flow rather than precipitation associated with tropical cyclones (TCs), we omit from our analysis all days during which the center of a TC was within 100 km of the Florida coastline. We specifically use data from the Atlantic hurricane database (HURDAT2; Landsea and Franklin 2013) to identify tropical depressions, tropical storms, and hurricanes that came within 100 km of the Florida coastline. This resulted in the removal of 170 days from the 5520 total days in the 1961–2020 period considered in this study, 92 from the 1961–90 climatological period and 78 from the 1991–2020 period.
b. Methods
Following the methodology of the original NWS WFO Tampa study, Mroczka (2021), and Chavez et al. (2022), we calculate sea-breeze regimes by averaging 1000–700-hPa vector wind at 1200 UTC each day during JJA for a region centered approximately over Tampa, Florida, from 82° to 83°W to 27.5° to 28.5°N. As Byers and Rodebush (1948) note that 1) the sea breeze typically does not develop until 2–3 h after sunrise and 2) the land breeze generally diminishes around sunrise, we assume that the winds at 1200 UTC, or 0800 local time, result primarily from synoptic-scale forcing during the Florida rainy season. While Mroczka (2021) initially defined the regimes on the basis of vector wind over Tampa given their interest in improving sea-breeze convection forecasts for the NWS WFO Tampa County warning area, they found that these regimes affected precipitation probabilities more broadly throughout the Florida peninsula.
Excluding days during which a TC existed within 100 km of the Florida coastline, we assign each day in JJA to one of the nine sea-breeze regimes defined in Table 1. Given the relative rarity of the regimes with flow from the north through northwest (hereinafter N/NW) (e.g., strong N/NW flow as in regime 9 only occurred on 18 days during the 1991–2020 period), we omit these regimes when assessing interannual variability and trends. While the true onset and demise of the Florida rainy season varies by year and location (Misra and DiNapoli 2013), we limit our analysis to the months of June, July, and August to capture the peak rainy season for most of the peninsula and avoid possible interference from more frequent nontropical and tropical disturbances in May and September, respectively.
Sea-breeze regime wind direction and speed thresholds. Values represent thresholds for the mean vector wind in the 1000–700-hPa layer calculated within a 1° by 1° box centered over Tampa. For direction ranges, a square bracket indicates that the end point is included and a parenthesis indicates that the end point is excluded. Note that, consistent with Mroczka (2021) and Chavez et al. (2022), we use units of knots (1 kt = 0.5144 m s−1).
3. Sea-breeze regime climatology: 1991–2020
a. Climatological sea-breeze regime distribution
The regime distribution during the 1991–2020 climatological period presented in Fig. 1 indicates that regime 4 (moderate flow from west through southwest; hereinafter W/SW), which accounts for almost a quarter of all regime days each month in JJA and as much as 26% in the month of July, occurs most frequently. Regime 1, which represents weak synoptic-scale flow with winds less than 4 kt (1 kt ≈ 0.51 m s−1) and accounts for about 15% of the JJA distribution, is the second most common regime during the most recent 1991–2020 climatological period.
The relative frequency (%) of each sea-breeze regime during the 1991–2020 period (green) and 1961–90 period (brown). (top) The JJA-season relative frequencies. (bottom) The relative frequencies for each individual month for each regime.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
Moderate flow from the east through northeast (hereinafter E/NE) in regime 2, which accounts for approximately 14% of days during JJA overall and as much as 16% during June, occurs about as frequently as moderate flow from the south through southeast (hereinafter S/SE) in regime 6. E/NE flow becomes less common during July when days with S/SE and W/SW flow increase in frequency. While this section focuses on the 1991–2020 climatological period, we note that Fig. 1 also includes the sea-breeze regime distribution over the 1961–90 period. In particular, Fig. 1 depicts apparent decreases in the relative frequencies of sea-breeze regimes with an easterly wind component and modest increases in the frequency of weak flow and W/SW flow, with the greatest differences occurring in August. These apparent changes in the sea-breeze regime climatology will be discussed in more detail in section 4.
Regime 9 (strong N/NW flow) occurs least frequently, accounting for less than 1% of days in JJA. The lack of regime-9 days is consistent with climatological JJA circulation patterns over the southern United States; strong N/NW flow over Florida typically requires deep troughing to extend well into the peninsula—a very rare occurrence during the summer months.
b. Sea-breeze regime composite analysis
We employ composite analysis to examine the mean atmospheric environment associated with each sea-breeze regime. All composite means are calculated for the 1991–2020 period. 850-hPa height composites (Fig. 2) also include averages for the 1961–91, 1971–2000, and 1981–2010 periods to highlight the westward expansion of the NASH noted in Li et al. (2011). We calculate standardized anomalies of 500-hPa geopotential height, total precipitable water, and 1000–850-hPa divergence by subtracting the daily climatologies for each day in JJA and dividing by the daily standard deviation before taking the composite mean for each regime.
Composite mean 850-hPa geopotential heights at 1200 UTC for each sea-breeze regime (solid contours). The dashed line represents the approximate location of the 850-hPa western ridge axis. Composites are created for four overlapping climatological periods, 1961–90 (lightest purple), 1971–2000, 1981–2010, and 1991–2020 (darkest purple).
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
1) 850-hPa and 500-hPa geopotential heights
Composite mean 850- and 500-hPa heights at 1200 UTC (Figs. 2 and 3, respectively) show how the average strength and position of the NASH and its western ridge vary by regime. Prior research (e.g., Li et al. 2011; Diem 2013; Li et al. 2012; Nieto Ferreira and Rickenbach 2020) found that the intersection of the 1560-gpm contour and the 850-hPa ridge axis serves as a useful estimate for the western boundary of the NASH western ridge.
Using this definition, we note that the most westward-extended composite mean 850-hPa ridge occurs during E/NE flow (regimes 2 and 3), when the 1560-gpm line is located west of the Mississippi River delta at a longitude of 93.4° and 94.4°W for regimes 2 and 3, respectively. As expected, the western ridge also shifts farthest north during days with E/NE flow, with the composite mean ridge axis located well north of Florida. Composite 500-hPa height means and anomalies suggest that these E/NE regimes tend to be associated with anomalous continental ridging over the eastern half of the United States (Fig. 3).
Composite mean 500-hPa geopotential height standardized anomalies (shaded) and means (contour lines) at 1200 UTC for each sea-breeze regime during the 1991–2020 climatological period.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
In contrast, S/SE flow (regimes 6 and 7) occur with a westward expansion of the NASH and associated ridging aloft. The composite mean 1560-gpm contour is located west of the Florida peninsula (∼86°W) and the 850-hPa ridge axis is draped over the Florida Panhandle or adjacent southern states during S/SE flow regimes. W/SW flow develops when the 850-hPa ridge axis shifts south over the extreme southern peninsula (regime 4) or Florida Keys (regime 5). This southward displacement of the subtropical ridge tends to be associated with anomalous troughing over the eastern United States and Gulf of Mexico, with a particularly amplified composite mean trough for strong W/SW flow (regime 5). Weak flow during regime 1 typically occurs when the ridge axis extends across the northern half of the peninsula at a mean latitude of 29°N. During the most recent 1991–2020 climatological period, the western edge of the ridge extended well west of the Florida peninsula (89.6°W) on regime-1 days, with weak positive 500-hPa geopotential height anomalies over the southern states.
2) Precipitation ingredients
Cloutier-Bisbee et al. (2019) applied an ingredients-based precipitation analysis to examine the relationship between heat waves and subsequent heavy precipitation events in Florida. They identified atmospheric variables describing sources of lift, moisture, and instability that contribute to the development of convective precipitation. Given our interest in the relationship between sea-breeze regimes and precipitation, we follow a similar approach here and examine composite values of 1000–850-hPa divergence (lift; Fig. 4), total precipitable water (moisture; Fig. 5), and the K index (instability; not shown). We select these variables on the basis of prior investigations of convective environments over Florida (e.g., Fuelberg and Biggar 1994; Cloutier-Bisbee et al. 2019; Chavez et al. 2022). These composites use ERA5 data from 1200 to 2300 UTC, approximate daylight hours during JJA, to capture the atmospheric environment leading up to and during the most convectively active period within a 24-h calendar day.
Composite mean 1000–850-hPa divergence standardized anomalies (shaded; s−1) and means (contour lines; s−1 × 10−5) averaged over 1200–2300 UTC for each sea-breeze regime during the 1991–2020 climatological period. Dashed contour lines represent negative mean values (i.e., convergence).
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
Composite mean total precipitable water standardized anomalies (shaded) and means (contour lines) averaged over 1200 to 2300 UTC for each sea-breeze regime during the 1991–2020 climatological period. Gray wind barbs indicate the composite mean 1000–700-hPa wind at 1200 UTC.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
Byers and Rodebush (1948) found that, given the lack of synoptic-scale forcing over Florida during JJA, the primary source of lift comes from near-surface convergence. Cloutier-Bisbee et al. (2019) noted that convergence in the 1000–850-hPa layer tends to be most predictive of precipitation development over Florida. Thus, we examine composite mean and anomalous 1000–850-hPa divergence for each of the nine sea-breeze regimes in Fig. 4. The divergence plots largely match our expectations based on prior research, with enhanced near-surface convergence (purple shades in Fig. 4) concentrated on the coast for which the synoptic-scale wind opposes the daytime sea-breeze front. W/SW flow during regimes 4 and 5 results in enhanced convergence over the eastern peninsula, while S/SE flow results in enhanced convergence over the western peninsula. Weak synoptic-scale flow (regime 1) tends to yield anomalous convergence over inland Florida. This supports the notion that weak synoptic flow allows the formation and inland penetration of both the west and east coast sea-breeze fronts, which converge somewhere over interior Florida. With the exception of a narrow strip of enhanced convergence along the immediate west coast, positive divergence anomalies occur over much of the peninsula during E/NE flow (regime 2 and 3).
Total precipitable water (TPW) represents the total column water vapor and has been shown to be predictive of convective precipitation, and extreme precipitation in particular, by numerous prior studies (e.g., Kunkel et al. 2020; Chavez et al. 2022; Kim et al. 2022). Furthermore, forecasters in Florida often use TPW as a key metric for assessing the presence of sufficient atmospheric moisture to support convection. Figure 5 illustrates that while climatological TPW values consistently exceed 40 mm for all regimes, E/NE synoptic-scale flow during regimes 2 and 3 produce the largest negative TPW anomalies across the peninsula. As discussed previously, E/NE flow tends to result from enhanced ridging over the entire southeast United States. Weak synoptic-scale winds during regime-1 days are associated with weak negative TPW anomalies over most of the peninsula. In contrast, regimes 5 and 7, which represent strong W/SW and S/SE flow respectively, tend to yield the highest mean values peninsula-wide. Chavez et al. (2022) suggest that this results from enhanced moisture advection from surrounding ocean water. Additionally, we note that stronger winds in these regimes likely enhance evaporation from warm water in the Gulf of Mexico and western Atlantic, whereas weak flow in regime 1 reduces evaporation and moisture advection.
Although S/SE flow in regimes 6 and 7 on average coincides with positive TPW anomalies over the western peninsula and Florida Panhandle, the eastern peninsula typically experiences larger TPW anomalies during W/SW flow in regimes 4 and 5. The greater westward extension of the NASH western ridge during regimes 6 and 7 (Figs. 2 and 3) results in a relatively drier air mass east of the Florida peninsula and may contribute to lower TPW values over the eastern peninsula. Interpreting Fig. 5 alongside Fig. 4, we can see that this pattern of enhanced TPW over the western peninsula during S/SE flow regimes may also result from greater moisture convergence as the synoptic-scale wind opposes the west-coast sea breeze. Likewise, the same pattern of enhanced TPW over the eastern peninsula during W/SW flow regimes may also result from greater moisture convergence as the synoptic-scale wind opposes the east-coast sea breeze.
3) Precipitation
The mean daily probability of precipitation (PoP) during JJA varies from greater than 70% over much of the southwestern peninsula to around 50%–60% over the western panhandle and extreme northern peninsula (Fig. 6). Here we define the probability of precipitation as the total number of JJA days in the 1991–2020 period during which more than 0.5-mm of rainfall occurred within a given grid box divided by the total number of JJA days without tropical cyclones in the 1991–2020 period. High PoP values of over 60% are common throughout the state during JJA, particularly given the 0.25° × 0.25° spatial resolution of the precipitation dataset used here.
Mean JJA daily probability of precipitation values for the 1991–2020 climatological period.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
Regimes 1, 2, 3, 6, and 7, which together account for over 50% of days in JJA during a typical summer (Fig. 1), all have local maxima in PoP values over the southwest or west-central portion of the peninsula (Fig. 7). This may help to explain the mean JJA PoP pattern in Fig. 6, with maximum season-mean PoP values over southwestern Florida. More than one-half of all summer days fall into a regime with either weak or easterly flow that enhances surface convergence and convective development over the western peninsula.
Regime composite JJA daily probability of precipitation values for the 1991–2020 climatological period.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
The spatial patterns of PoP values for each regime (Fig. 7) align well with the geopotential height and precipitation ingredient composites provided in Figs. 2–5. With the exception of the southern peninsula, the lowest probabilities occur during E/NE flow (regimes 2 and 3) when anomalous ridging, negative TPW anomalies, and positive 1000–850-hPa divergence anomalies (weaker near-surface convergence) extend over much of the Florida peninsula. PoP values also fall below 60% during N/NW flow, particularly over the panhandle and southwestern peninsula. The highest rainfall probabilities generally occur on days with strong W/SW or S/SE flow, which highlights the key influence of moisture advection in precipitation coverage during JJA. Strong W/SW flow results in PoP values exceeding 90% across much of the eastern peninsula, and strong S/SE flow results in PoP values exceeding 90% across much of the western peninsula. The presence of a trough over the eastern United States during regime 5 (strong W/SW flow) may also contribute upper-level dynamic support for enhanced precipitation over the peninsula.
4. Trends in the sea-breeze regime climatology: 1961–2020
A key goal of this research is to identify whether the climatological sea-breeze regime distribution has shifted in response to changes in the strength and extent of the NASH and its western ridge. The regime relative frequency distributions in Fig. 1 suggest that over the two climatological periods, 1961–90 and 1991–2020, some regimes have increased in frequency (e.g., regimes 1 and 4), while others have become less common (e.g., regimes 2 and 6). In this section we employ a more thorough analysis to explore the significance of these changes. Given the relative infrequency of N/NW flow (regimes 8 and 9), we omit these two regimes from this analysis.
Figure 8 depicts the annual relative frequency of each regime during JJA (blue bars) and also provides an estimate of the 1961–2020 monotonic trend (black lines) obtained using a simple Mann–Kendall trend test (Mann 1945; Kendall 1948). In addition to substantial interannual variability, Fig. 8 indicates an increase in the relative frequency of regimes 1 (weak flow) and 4 (moderate W/SW flow) with corresponding decreases in regimes 2 (moderate E/NE flow) and 6 (moderate S/SE flow). We test the sensitivity of these trends to changes in the starting year by calculating the trend for starting years varying from 1961 to 1991 with the ending year held at 2020 (Fig. 9). The trend values depicted in Fig. 9 represent the Sen’s slope parameter from a Mann–Kendall test that uses the prewhitening algorithm proposed in Collaud Coen et al. (2020) to simultaneously maximize the power of the test while minimizing the likelihood of type 1 errors. Viewed with the 90% and 95% confidence intervals, very few of the apparent trends in sea-breeze regime relative frequency appear significant. Regime 4 (moderate W/SW) exhibits a significant positive trend of approximately +1%–1.5% per decade for start dates from 1961 through 1978 with more modest trends of about +1% per decade after 1978. The positive trend in regime 1 (weak flow) does not appear to be significant at either the 90% or 95% confidence level. In contrast, regime 6 (moderate S/SE flow) maintains a weak negative trend of roughly −1% per decade for most of the period. Interestingly, a negative trend in regime 2 (moderate N/NE flow) increases in magnitude throughout the period, reaching values of −2% to −3% per decade for starting dates after the late 1970s.
The relative frequency (%) of each sea-breeze regime per year during the period 1961–2020. The solid black lines represent the trend (Sen’s slope) resulting from a simple Mann–Kendall trend test.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
Trends in the annual relative frequency of each sea-breeze regime for start dates ranging from 1961 to 1991 and a constant end date of 2020. The y axis gives the trend value, and the x axis gives the start date. The light- and dark-blue-shaded areas represent the 95% and 90% confidence intervals, respectively. Trend values represent the Sen’s slope parameter of a nonparametric Mann–Kendall trend test with prewhitening.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
All else equal, we might expect the westward shift in the climatological position of the NASH western ridge as reported in Li et al. (2011) and confirmed by us (not shown) to result from an increase in the prevalence of sea-breeze regimes 2 and 3 (E/NE flow) or perhaps regimes 6 and 7 (S/SE flow), given that the western ridge tends to be displaced farther west during these four regimes (Fig. 2). As indicated in Figs. 7 and 8, this does not appear to be the case. Instead, regime 4, a regime that features a more eastward-retracted NASH and western ridge, has increased in frequency while regimes 2, 3, 6, and 7 have either maintained or decreased in frequency.
The westward expansion of the climatological NASH and its western ridge may instead result from a westward shift in the composite ridge position during some or all of the sea-breeze regimes. Indeed, a closer examination of Fig. 2 indicates that the composite NASH western ridge, as denoted by the 1560-gpm contour, has experienced a substantial 11.3° westward migration in regime 1 (weak flow) and a more modest 3.5° westward migration in regime 4 (W/SW moderate) when comparing the mean positions between the 1961–90 and 1991–2020 periods. During the climatological periods 1961–90 and 1971–2000, the NASH western ridge is located east of the Florida peninsula at a longitude of 78.4°W during regime 1. By the 1991–2020 period, the regime-1 western ridge extends west of the Florida Panhandle to a longitude of 89.7°W. The regime-4 western ridge has shifted from the Bahamas (78.1°W) to the southwest coast of the Florida peninsula (81.6°W). Except for subtle shifts in the 1560-gpm contour for regimes 8 and 9, westward migration of the NASH ridge appears to be confined to regimes 1 and 4. In other words, while we do not observe an increase in the frequency of regimes with a westward-expanded ridge (regimes 2, 3, 6, and 7), we do see a westward shift in the composite ridge position for regimes that have increased in frequency and that together make up the largest share of the regime distribution (regimes 1 and 4). This result aligns well with the thermodynamic mechanism of western ridge expansion proposed in Li et al. (2012), which they noted tends to yield a more southwestward-shifted ridge.
Last, we note that, when comparing the 1991–2020 and 1961–90 climatological periods, Fig. 1 suggests that changes in the sea-breeze regime distribution have occurred primarily later in the summer. With the exception of regime 2 (moderate E/NE flow), the regime relative frequencies did not change by more than 1% during June. In contrast, by August, regimes 1 and 4 increased by 5% and 4%, respectively, while regimes 2 and 6 decreased by 6% and 5%, respectively. On average, the NASH intensifies and expands westward as the summer progresses, particularly in late July and August (Gamble et al. 2008). Thus, the observed changes in the sea-breeze regime distribution occur closer to the climatological maximum of the NASH rather than earlier in the summer. This similarly has implications for precipitation over the Caribbean region. In general, precipitation tends to be enhanced on the northern side of the western ridge and suppressed on the southern side (Li et al. 2012); thus, a westward-expanded ridge centered over the south Florida peninsula during increasingly common regime-4 days may yield drier conditions over parts of the western Caribbean.
5. The relationship between interannual variability in sea-breeze regimes and precipitation
In addition to possible trends, Fig. 8 indicates the presence of substantial interannual variability in the sea-breeze regime distribution. For example, the relative frequency of moderate W/SW flow varies from a minimum of 10% in 1984 to a maximum of 37% in 2014, while the relative frequency of moderate S/SE flow varies from 4% in 1995 to 27% in 1965 and 2003. Interannual variability in the mean summer position and strength of the NASH has been well documented by numerous prior studies, many of which linked variability in the NASH with variability in precipitation or heat stress over the southeastern United States (e.g., Mächel et al. 1998; Hasanean 2004; Li et al. 2011, 2019; Luo et al. 2021). We note that the causes of interannual variability in the NASH remain debated, with suggested mechanisms including the Atlantic multidecadal oscillation, the Pacific decadal oscillation, Rossby wave trains originating over the North Pacific, and teleconnections arising from the South Asian monsoon (e.g., Kushnir et al. 2010; Hu et al. 2011; Kelly and Mapes 2011; Li et al. 2012; Luo et al. 2021).
We have already demonstrated the influence of sea-breeze regimes on the spatial patterns of daily precipitation probabilities (Fig. 7). In this section we examine the relationship between trends and interannual variability in sea-breeze regime frequency and trends and interannual variability in precipitation probabilities. Combining the results presented so far, we might expect the observed increases or decreases in the relative frequency of moderate W/SW or E/NE flow, respectively, to be associated with increases in the frequency of days with precipitation over the eastern half of the Florida peninsula. Similarly, we might expect a decrease in the number of days with precipitation over the western peninsula given the decline in E/NE and S/SE flow. To examine changes in precipitation frequency, we compute the annual relative frequency of “rainy days” over the 1961–2020 period, where a “rainy day” for a given grid box is defined as daily precipitation values exceeding 0.5-mm. Because we remove days with TCs within 100-km of the Florida coastline from our dataset, we divide the total number of rainy days for a given year by the total number of TC-free JJA days that year. We apply a Mann–Kendall trend test to the annual rainy-day frequency data and test for field significance using the false discovery rate (Wilks 2006, 2016).
Most of the statistically significant trends in rainy-day relative frequency occur at grid boxes along the immediate coast (Fig. 10). In particular, we observe statistically significant positive rainy-day frequency trends of greater than 2.4% per decade for areas along the east coast of Florida, north of Cape Canaveral. Smaller but still significant positive trends exist along much of the southwest Florida coastline and for portions of the Florida Panhandle. The spatial pattern of trends along the coastline mostly aligns with prior research. For example, Martinez et al. (2012) examined station data and found that significant positive precipitation trends during JJA for the period 1895–2009 were limited to coastal locations across south Florida, though we note that their analysis considered season total precipitation rather than rainy-day frequency and did not explore possible explanations for the observed trends.
Trends in the annual relative frequency of days with precipitation exceeding 0.5 mm (% yr−1). Trend values represent the Sen’s slope parameter of a nonparametric Mann–Kendall trend test. Grid boxes outlined in yellow represent field significant trends at the 5% significance level.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
The positive trends in rainy-day frequency along the northeast Florida coast overall match our expectations based on the increased frequency of moderate W/SW flow, though it is worth noting that the largest precipitation probabilities for moderate W/SW flow occur just inland of the coast (Fig. 7). In contrast, the positive trends along the west coast of Florida are less expected given the apparent decline in easterly flow days.
To further examine the relationship between sea-breeze regimes and precipitation variability on an interannual time scale, we calculate the Pearson product-moment correlation between the annual regime relative frequency and the annual rainy-day relative frequency at each grid point (Fig. 11). Figure 11 depicts correlation maps for regimes 2–7, the only regimes significantly correlated to precipitation variability. Figure 11 indicates a significant negative correlation between the annual relative frequency of regimes 2 and 3 (E/NE flow) and annual rainy-day frequency over much of the northern peninsula, particularly along the northeast coast. Anomalous ridging over the southeastern United States and reduced precipitable water contribute to drier conditions and a lower probability of precipitation over northern and northeastern Florida during E/NE flow. Thus, summers with a greater prevalence of E/NE flow tend to be associated with fewer rainy days. As expected, moderate W/SW flow is positively correlated to rainy-day frequency along the northeastern coast of the peninsula and over portions of central Florida, while moderate S/SE flow is positively correlated to rainy-day frequency over portions of the Florida Panhandle and a small stretch of the west coast of the peninsula.
Pearson product-moment correlation coefficient between annual regime relative frequency and annual rainy-day relative frequency. Thick black outlines indicate field significance at the 5% significance level.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
Figure 11 also suggests that interannual variability in the sea-breeze regime distribution is unrelated to interannual variability in rainy-day frequency over south Florida. No significant correlations exist between regime relative frequency and rainy-day frequency for any of the regimes over the extreme southern peninsula. Other factors, such as sea surface temperature anomalies or atmospheric moisture, may play a larger role in modulating rainy-day frequency over south Florida. Surface convergence and convective development over the southern peninsula are further complicated by interactions between sea-breeze boundaries all along the southern tip of the peninsula and lake-breeze boundaries associated with Lake Okeechobee (e.g., Boybeyi and Raman 1992). Additional complexity in sea-breeze processes may arise as a result of the sharp gradient in land surface type between the highly developed southeast coast and the sparsely populated Everglades region (e.g., Marshall et al. 2004). Despite these complicating factors over the southeastern Florida peninsula, it also is possible that modifications to the sea-breeze regime definitions, for example using the position of the 850-hPa western ridge rather than vector wind near Tampa or using vector wind over another region of the state, may yield regimes that better account for interannual variability in south Florida precipitation probabilities.
Although we primarily have focused on precipitation probabilities and rainy-day frequency, we similarly find that the sea-breeze regime distribution looks different for particularly rainy versus dry summers. First, we calculate the statewide-averaged total rainfall during JJA for each year in the 1961–2020 period and identify the 20 years with the lowest season-total rainfall (driest tercile) and the 20 years with the highest season-total rainfall (rainiest tercile). As illustrated in Fig. 12, the sea-breeze regime distributions for the driest tercile and rainiest tercile years differ. During drier summers, the relative frequency of moderate E/NE flow is approximately 4% higher than during rainier summers. In contrast, rainy summers feature 4% more days with moderate S/SE flow relative to dry summers. These results are not particularly surprising; summers with more southerly flow, particularly from the Caribbean, result in higher season-total precipitation, while summers with more northerly flow result in lower season-total precipitation. We previously emphasized the role of sea-breeze regimes in modulating spatial variations in precipitation probabilities within the state, but these results suggest that interannual variability in the sea-breeze regime distribution also impacts statewide precipitation totals during the summer months.
Regime relative frequencies for the rainiest (green bars) vs the driest (brown bars) 20 years in the 1961–2020 period. The rainiest and driest years respectively represent the top and bottom tercile, or 20 years, of statewide-averaged JJA season-total rainfall, excluding days on which a TC was within 100 km of the Florida coastline.
Citation: Journal of Applied Meteorology and Climatology 63, 2; 10.1175/JAMC-D-23-0151.1
6. Summary and discussion
The broad goals of this study were to examine the full 30-yr Florida sea-breeze regime climatology, the long-term variability and trends in the sea-breeze regime distribution, and the impacts of such variability and trends on interannual variability and trends in precipitation occurrence. The full 30-yr sea-breeze regime climatology indicates that moderate W/SW flow occurs most frequently (roughly 25% of JJA days on average), followed by weak flow (winds < 4 kt), moderate E/NE flow, and moderate S/SE flow. N/NW flow and winds of any direction exceeding 10 kt occur much less frequently. W/SW flow on average is associated with a southward shift in the western ridge axis and anomalous troughing over eastern United States, while S/SE and E/NE flow tend to develop when the western ridge shifts northward and westward, with E/NE flow in particular also coinciding with enhanced continental ridging over the southeastern United States. Regimes with wind speeds exceeding 10 kt yield higher total precipitable water values statewide and generally coincide with the highest precipitation probabilities. As expected, W/SW flow enhances surface convergence along the east coast sea-breeze front and leads to higher precipitation probabilities along the eastern half of the peninsula, whereas S/SE flow enhances surface convergence along the west coast sea-breeze front and leads to higher precipitation probabilities along the western half of the peninsula. Lower precipitation probabilities over much of the peninsula, though especially north of the Everglades, occur on days with E/NE flow when enhanced subsidence from continental ridging and negative precipitable water anomalies dominate.
Li et al. (2011, 2012) found that the climatological position of the NASH western ridge has shifted westward in time. Given that the nine sea-breeze regimes are closely related to the position of the NASH western ridge, we also examined interannual variability and trends in the relative frequency of each regime over the 1961–2020 period. A nonparametric trend analysis suggests an increase in the number of days per summer under regime 4 (moderate W/SW flow) and a decrease in the relative frequency of regime 2 (moderate E/NE flow). Beyond trends, we find substantial interannual variability in the sea-breeze regime distribution, with some summers experiencing large deviations from the climatological distribution. Furthermore, we find significant correlations between interannual variability in the relative frequency of some regimes and interannual variability in the relative frequency of rainy days for some parts of the Florida peninsula. In particular, the northeastern portion of the peninsula tends to experience more rainy days during summers with a higher relative frequency of W/SW flow. In contrast, sea-breeze regime frequency is not significantly correlated with rainy-day frequency across the southern Florida peninsula. This suggests the need to define new sea-breeze regimes, perhaps based on the location of the western ridge or using mean winds averaged over a different region of the peninsula, and the potential importance of other variables such as sea surface temperature anomalies, midlevel lapse rates, and atmospheric moisture in modulating summertime precipitation over the southern peninsula.
Limitations of this study include our lack of attention to additional factors such as soil moisture that are known to influence sea-breeze convergence and convective initiation along the sea-breeze front (Baker et al. 2001). We also do not consider the role of sea surface temperature anomalies in the Gulf of Mexico, Caribbean Sea, and western North Atlantic Ocean, which have been shown to affect interannual variability in warm-season rainfall over Florida and the southeastern United States (e.g., Kushnir et al. 2010; Hu et al. 2011; Misra and Mishra 2016). Also, although we remove TCs from our analysis, other synoptic-scale disturbances such as easterly waves and the rare frontal passage may muddy the results.
While Florida’s sea-breeze regimes primarily have been used for weather prediction applications, our research demonstrates the potential utility of sea-breeze regimes for climate prediction applications, though we acknowledge that additional predictors such as sea surface temperature will likely be necessary. Although prior research suggests limited predictability of the NASH on seasonal time scales (e.g., Stefanova et al. 2012; Infanti and Kirtman 2014), future work might employ model ensembles such as the Subseasonal Experiment (Pegion et al. 2019) to examine the predictability of the NASH and related sea-breeze regimes on subseasonal time scales. Additional future work will explore alternative sea-breeze regimes that incorporate other variables such as sea surface temperature anomalies and midlevel atmospheric moisture and lapse rates.
Acknowledgments.
We acknowledge and appreciate constructive feedback provided by Dan Halperin and two anonymous reviewers.
Data availability statement.
All data used in this study come from publicly available datasets. ERA5 data (Hersbach et al. 2020) were downloaded from the Copernicus Climate Change Service (C3S; https://doi.org/10.24381/cds.bd0915c6). CPC Unified Gauge-Based Precipitation data (Xie et al. 2007; Chen et al. 2008) were obtained from the Columbia Climate School’s International Research Institute for Climate and Society (IRI) online Climate Data Library. Tropical cyclone tracks within 100 km of the Florida coastline from the HURDAT2 dataset (Landsea and Franklin 2013) were obtained online (https://coast.noaa.gov/hurricanes/).
REFERENCES
Arritt, R. W., 1993: Effects of the large-scale flow on characteristic features of the sea breeze. J. Appl. Meteor., 32, 116–125, https://doi.org/10.1175/1520-0450(1993)032<0116:EOTLSF>2.0.CO;2.
Baker, R. D., B. H. Lynn, A. Boone, W.-K. Tao, and J. Simpson, 2001: The influence of soil moisture, coastline curvature, and land-breeze circulations on sea-breeze-initiated precipitation. J. Hydrometeor., 2, 193–211, https://doi.org/10.1175/1525-7541(2001)002<0193:TIOSMC>2.0.CO;2.
Boybeyi, Z., and S. Raman, 1992: A three-dimensional numerical sensitivity study of convection over the Florida peninsula. Bound.-Layer Meteor., 60, 325–359, https://doi.org/10.1007/BF00155201.
Burpee, R. W., and L. N. Lahiffi, 1984: Area-average rainfall variations on sea-breeze days in south Florida. Mon. Wea. Rev., 112, 520–534, https://doi.org/10.1175/1520-0493(1984)112<0520:AARVOS>2.0.CO;2.
Byers, H. R., and H. R. Rodebush, 1948: Causes of thunderstorms of the Florida peninsula. J. Atmos. Sci., 5, 275–280, https://doi.org/10.1175/1520-0469(1948)005<0275:COTOTF>2.0.CO;2.
Chavez, I., S. M. Milrad, D. J. Halperin, B. Mroczka, and K. R. Tyle, 2022: Environmental analysis of warm-season first cloud-to-ground lightning events over the western Florida peninsula. Wea. Forecasting, 37, 1867–1883, https://doi.org/10.1175/WAF-D-22-0005.1.
Chen, M., W. Shi, P. Xie, V. B. S. Silva, V. E. Kousky, R. Wayne Higgins, and J. E. Janowiak, 2008: Assessing objective techniques for gauge-based analyses of global daily precipitation. J. Geophys. Res., 113, D04110, https://doi.org/10.1029/2007JD009132.
Cloutier-Bisbee, S. R., A. Raghavendra, and S. M. Milrad, 2019: Heat waves in Florida: Climatology, trends, and related precipitation events. J. Appl. Meteor. Climatol., 58, 447–466, https://doi.org/10.1175/JAMC-D-18-0165.1.
Collaud Coen, M., E. Andrews, A. Bigi, G. Martucci, G. Romanens, F. P. A. Vogt, and L. Vuilleumier, 2020: Effects of the prewhitening method, the time granularity, and the time segmentation on the Mann–Kendall trend detection and the associated Sen’s slope. Atmos. Meas. Tech., 13, 6945–6964, https://doi.org/10.5194/amt-13-6945-2020.
Diem, J. E., 2013: Influences of the Bermuda high and atmospheric moistening on changes in summer rainfall in the Atlanta, Georgia region, USA. Int. J. Climatol., 33, 160–172, https://doi.org/10.1002/joc.3421.
Duever, M. J., J. F. Meeder, L. C. Meeder, and J. M. McCollom, 1994: The climate of south Florida and its role in shaping the everglades ecosystem. Everglades: The Ecosystem and its Restoration, CRC Press, 225–248.
Fuelberg, H. E., and D. G. Biggar, 1994: The preconvective environment of summer thunderstorms over the Florida Panhandle. Wea. Forecasting, 9, 316–326, https://doi.org/10.1175/1520-0434(1994)009<0316:TPEOST>2.0.CO;2.
Gamble, D. W., D. B. Parnell, and S. Curtis, 2008: Spatial variability of the Caribbean mid-summer drought and relation to North Atlantic high circulation. Int. J. Climatol., 28, 343–350, https://doi.org/10.1002/joc.1600.
Gentry, R. C., and P. L. Moore, 1954: Relation of local and general wind interaction near the sea coast to time and location of air-mass showers. J. Atmos. Sci., 11, 507–511, https://doi.org/10.1175/1520-0469(1954)011<0507:ROLAGW>2.0.CO;2.
Hasanean, H. M., 2004: Variability of the North Atlantic subtropical high and associations with tropical sea-surface temperature. Int. J. Climatol., 24, 945–957, https://doi.org/10.1002/joc.1042.
Hersbach, H., and Coauthors, 2020: The ERA5 global reanalysis. Quart. J. Roy. Meteor. Soc., 146, 1999–2049, https://doi.org/10.1002/qj.3803.
Hu, Q., S. Feng, and R. J. Oglesby, 2011: Variations in North American summer precipitation driven by the Atlantic multidecadal oscillation. J. Climate, 24, 5555–5570, https://doi.org/10.1175/2011JCLI4060.1.
Infanti, J. M., and B. P. Kirtman, 2014: Southeastern U.S. rainfall prediction in the North American Multi-Model Ensemble. J. Hydrometeor., 15, 529–550, https://doi.org/10.1175/JHM-D-13-072.1.
Kelly, P., and B. Mapes, 2011: Zonal mean wind, the Indian monsoon, and July drying in the western Atlantic subtropics. J. Geophys. Res., 116, D00Q07, https://doi.org/10.1029/2010JD015405.
Kendall, M. G., 1948: Rank Correlation Methods. Griffin, 160 pp.
Kim, S., A. Sharma, C. Wasko, and R. Nathan, 2022: Linking total precipitable water to precipitation extremes globally. Earth’s Future, 10, e2021EF002473, https://doi.org/10.1029/2021EF002473.
Klavans, J. M., A. C. Clement, L. N. Murphy, and H. Zhang, 2020: Identifying the externally forced Atlantic multidecadal variability signal through Florida rainfall. Geophys. Res. Lett., 47, e2020GL088361, https://doi.org/10.1029/2020GL088361.
Kunkel, K. E., S. E. Stevens, L. E. Stevens, and T. R. Karl, 2020: Observed climatological relationships of extreme daily precipitation events with precipitable water and vertical velocity in the contiguous United States. Geophys. Res. Lett., 47, e2019GL086721, https://doi.org/10.1029/2019GL086721.
Kushnir, Y., R. Seager, M. Ting, N. Naik, and J. Nakamura, 2010: Mechanisms of tropical Atlantic SST influence on North American precipitation variability. J. Climate, 23, 5610–5628, https://doi.org/10.1175/2010JCLI3172.1.
Landsea, C. W., and J. L. Franklin, 2013: Atlantic hurricane database uncertainty and presentation of a new database format. Mon. Wea. Rev., 141, 3576–3592, https://doi.org/10.1175/MWR-D-12-00254.1.
Li, L., W. Li, and Y. Kushnir, 2012: Variation of the North Atlantic subtropical high western ridge and its implication to southeastern US summer precipitation. Climate Dyn., 39, 1401–1412, https://doi.org/10.1007/s00382-011-1214-y.
Li, W., L. Li, R. Fu, Y. Deng, and H. Wang, 2011: Changes to the North Atlantic subtropical high and its role in the intensification of summer rainfall variability in the southeastern United States. J. Climate, 24, 1499–1506, https://doi.org/10.1175/2010JCLI3829.1.
Li, W., T. Zou, L. Li, Y. Deng, V. T. Sun, Q. Zhang, J. B. Layton, and S. Setoguchi, 2019: Impacts of the North Atlantic subtropical high on interannual variation of summertime heat stress over the conterminous United States. Climate Dyn., 53, 3345–3359, https://doi.org/10.1007/s00382-019-04708-1.
Luo, H., Á. F. Adames, and R. B. Rood, 2021: A Northern Hemispheric wave train associated with interannual variations in the Bermuda high during boreal summer. J. Climate, 34, 6163–6173, https://doi.org/10.1175/JCLI-D-20-0608.1.
Mächel, H., A. Kapala, and H. Flohn, 1998: Behaviour of the centres of action above the Atlantic since 1881. Part I: Characteristics of seasonal and interannual variability. Int. J. Climatol., 18, 1–22, https://doi.org/10.1002/(SICI)1097-0088(199801)18:1%3C1::AID-JOC225%3E3.0.CO;2-A.
Mann, H. B., 1945: Nonparametric tests against trend. Econometrica, 13, 245–259, https://doi.org/10.2307/1907187.
Marshall, C. H., R. A. Pielke Sr., L. T. Steyaert, and D. A. Willard, 2004: The impact of anthropogenic land-cover change on the Florida peninsula sea breezes and warm season sensible weather. Mon. Wea. Rev., 132, 28–52, https://doi.org/10.1175/1520-0493(2004)132<0028:TIOALC>2.0.CO;2.
Martinez, C. J., J. J. Maleski, and M. F. Miller, 2012: Trends in precipitation and temperature in Florida, USA. J. Hydrol., 452–453, 259–281, https://doi.org/10.1016/j.jhydrol.2012.05.066.
Misra, V., and S. M. DiNapoli, 2013: Understanding the wet season variations over Florida. Climate Dyn., 40, 1361–1372, https://doi.org/10.1007/s00382-012-1382-4.
Misra, V., and A. Mishra, 2016: The oceanic influence on the rainy season of peninsular Florida. J. Geophys. Res. Atmos., 121, 7691–7709, https://doi.org/10.1002/2016JD024824.
Misra, V., L. Moeller, L. Stefanova, S. Chan, J. J. O’Brien, T. J. Smith III, and N. Plant, 2011: The influence of the Atlantic warm pool on the Florida Panhandle sea breeze. J. Geophys. Res., 116, D00Q06, https://doi.org/10.1029/2010JD015367.
Mroczka, B., 2021: Florida summer thunderstorm website. User’s guide, 7 pp., https://www.weather.gov/media/tbw/seabreeze/User_Guide_SB.pdf.
Nieto Ferreira, R., and T. M. Rickenbach, 2020: Effects of the North Atlantic subtropical high on summertime precipitation organization in the southeast United States. Int. J. Climatol., 40, 5987–6001, https://doi.org/10.1002/joc.6561.
Obeysekera, J., J. Browder, L. Hornung, and M. A. Harwell, 1999: The natural south Florida system I: Climate, geology, and hydrology. Urban Ecosyst., 3, 223–244, https://doi.org/10.1023/A:1009552500448.
Pegion, K., and Coauthors, 2019: The Subseasonal Experiment (SubX): A multimodel subseasonal prediction experiment. Bull. Amer. Meteor. Soc., 100, 2043–2060, https://doi.org/10.1175/BAMS-D-18-0270.1.
Price, R. M., and P. K. Swart, 2006: Geochemical indicators of groundwater recharge in the surficial aquifer system, Everglades National Park, Florida, USA. Perspectives on Karst Geomorphology, Hydrology, and Geochemistry, R. S. Harmon and C. Wicks, Eds., Geological Society of America, 251–266, https://doi.org/10.1130/2006.2404(21).
Roeder, W. P., B. H. Cummins, K. L. Cummins, R. L. Holle, and W. S. Ashley, 2015: Lightning fatality risk map of the contiguous United States. Nat. Hazards, 79, 1681–1692, https://doi.org/10.1007/s11069-015-1920-6.
Stefanova, L., V. Misra, J. J. O’Brien, E. P. Chassignet, and S. Hameed, 2012: Hindcast skill and predictability for precipitation and two-meter air temperature anomalies in global circulation models over the southeast United States. Climate Dyn., 38, 161–173, https://doi.org/10.1007/s00382-010-0988-7.
Wei, W., W. Li, Y. Deng, and S. Yang, 2019: Intraseasonal variation of the summer rainfall over the southeastern United States. Climate Dyn., 53, 1171–1183, https://doi.org/10.1007/s00382-018-4345-6.
Wilks, D. S., 2006: On “field significance” and the false discovery rate. J. Appl. Meteor. Climatol., 45, 1181–1189, https://doi.org/10.1175/JAM2404.1.
Wilks, D. S., 2016: “The stippling shows statistically significant grid points”: How research results are routinely overstated and overinterpreted, and what to do about it. Bull. Amer. Meteor. Soc., 97, 2263–2273, https://doi.org/10.1175/BAMS-D-15-00267.1.
Xie, P., M. Chen, S. Yang, A. Yatagai, T. Hayasaka, Y. Fukushima, and C. Liu, 2007: A gauge-based analysis of daily precipitation over East Asia. J. Hydrometeor., 8, 607–626, https://doi.org/10.1175/JHM583.1.
Zorzetto, E., and L. Li, 2021: Impacts of the North Atlantic subtropical high on daily summer precipitation over the conterminous United States. J. Hydrometeor., 22, 1697–1712, https://doi.org/10.1175/JHM-D-20-0242.1.
