Abstract

This study contributes to the understanding of the relationship between air temperature and convection by analyzing the characteristics of rainfall at the storm and convective rain cell scales. High spatial–temporal resolution (1 km, 5 min) estimates from a uniquely long weather radar record (24 years) were coupled with near-surface air temperature over Mediterranean and semiarid regions in the eastern Mediterranean. In the examined temperature range (5°–25°C), the peak intensity of individual convective rain cells was found to increase with temperature, but at a lower rate than the 7%°C−1 scaling expected from the Clausius–Clapeyron relation, while the area of the individual convective rain cells slightly decreases or, at most, remains unchanged. At the storm scale, the areal convective rainfall was found to increase with warmer temperatures, whereas the areal nonconvective rainfall and the stormwide area decrease. This suggests an enhanced moisture convergence from the stormwide extent toward the convective rain cells. Results indicate a reduction in the total rainfall amounts and an increased heterogeneity of the spatial structure of the storm rainfall for temperatures increasing up to 25°C. Thermodynamic conditions, analyzed using convective available potential energy, were determined to be similar between Mediterranean and semiarid regions. Limitations in the atmospheric moisture availability when shifting from Mediterranean to semiarid climates were detected and explain the suppression of the intensity of the convective rain cells when moving toward drier regions. The relationships obtained in this study are relevant for nearby regions characterized by Mediterranean and semiarid climates.

1. Introduction

Air temperature influences extreme rainfall intensity because of the increased water vapor–holding capacity of warmer air that follows the well-known Clausius–Clapeyron (CC) relation (e.g., Trenberth et al. 2003; O’Gorman and Schneider 2009; Fatichi et al. 2015). Convective rainfall is found to be particularly sensitive to warmer temperatures (Berg et al. 2013), as it is influenced by both thermodynamic and dynamic processes (Loriaux et al. 2013). The possible intensification of extreme rainfall intensity in a warmer climate has stimulated studies that quantify the relationship between temperature and rainfall rates [see review by Westra et al. (2014)]. A special focus in these studies is given to the scaling of extreme rainfall intensity with temperature, defined as the relative change of rainfall intensity in high quantiles per degree of increase in the air temperature. Daily to hourly scales were analyzed using observations and large-scale climate models for present (e.g., Eltahir and Pal 1996; Berg et al. 2009; Mishra et al. 2012; Fischer and Knutti 2016) and future climates (e.g., Ban et al. 2015; Bao et al. 2017). Furthermore, high-resolution climate models were used to study the response of rain intensity to near-surface temperature at fine spatiotemporal scales (e.g., Singleton and Toumi 2013; Meredith et al. 2015; Moseley et al. 2016; Haerter et al. 2017; Prein et al. 2017a). The scaling was observed to depend on rainfall type (i.e., convective versus stratiform; e.g., Moseley et al. 2013; Molnar et al. 2015), temporal resolution of the sample (e.g., Panthou et al. 2014), and location/climate (Drobinski et al. 2018; Pfahl et al. 2017).

So far, few contributions have targeted observation-based analyses of the impact of air temperature on the spatiotemporal characteristics of extreme rainfall (e.g., Berg et al. 2013; Wasko et al. 2015, 2016a,b). These studies focused on relatively large (continental) scales, using data derived from rain gauge networks or coarse-resolution remote sensing products. Moseley et al. (2013) and Lochbihler et al. (2017) analyzed the relationship between temperature or dewpoint temperature and the spatiotemporal characteristics of extreme rainfall at small scales using weather radar data. Yet, the effect of temperature on the small-scale structure of rainfall fields remains largely unexplored in many regions worldwide.

This study contributes to the understanding of the relationship between air temperature and convective rainfall by analyzing the characteristics of rainfall at the storm scale and at the scale of scattered convective elements, more specifically, individual convective rain cells. The study is focused on Mediterranean and semiarid climates, where convective rainfall largely contributes to the total annual rainfall (Sharon and Kutiel 1986). Convective rain cells often precipitate with a high rainfall intensity for a short duration (e.g., Peleg and Morin 2012) and are associated with the triggering of high-magnitude flash floods, especially in small- to medium-sized catchments in semiarid and arid climates (e.g., Syed et al. 2003; Morin et al. 2006; Yakir and Morin 2011). Their analysis requires high-resolution rainfall data in space and time that can be obtained using weather radars (e.g., Barnolas et al. 2010; Belachsen et al. 2017). Here we present how high-resolution rainfall data can be used for analyzing the scaling of rainfall properties with temperature to answer the following questions: How does temperature affect the intensification/weakening of convective rain cells? Is there an enhancement of moisture convergence and therefore cell intensity with temperature? If so, how does it affect the stormwide areal rainfall? How does the scaling change when shifting toward drier climates and along the sea–land transition?

A unique long-term record of high-resolution radar-based rainfall estimates was used to identify the convective rain cells over the eastern Mediterranean, a region characterized by different climatic conditions (from Mediterranean to arid), moisture availability, and atmospheric processes (thermodynamic and dynamic conditions). The scaling of peak rain intensity and area of convective rain cells with temperature was quantified. The relation between stormwide properties of convective rainfall fields and temperature was then estimated, paying attention to the role of the rainfall spatial structure at the storm scale.

2. Study area

The study focuses on the eastern Mediterranean (EM), a region characterized by a sharp change from Mediterranean to semiarid and arid climates (Fig. 1). Precipitation over the sea and the coastal region is generally brought by cold fronts and postfrontal systems associated with Mediterranean extratropical cyclones (Goldreich 2012), while precipitation over the semiarid and arid regions is brought by extratropical cyclones, Syrian lows, active Red Sea trough systems, and subtropical jet disturbances (Kahana et al. 2002; Goldreich 2012). While the mean annual precipitation decreases moving toward arid climates, the climatology of short-duration extreme rainfall, which is generally associated with convective rainfall, shows an increase of extreme intensities going toward drier areas (Morin et al. 2009; Marra and Morin 2015; Marra et al. 2017).

Fig. 1.

Map showing the weather radar (gray star), climatic zones (Mediterranean, green; semiarid, red; arid, orange) and the four 60 km × 60 km locations that were studied (square areas, labeled A–D). The meteorological stations of Ein Hahoresh, Beer Sheva, and Sedom are marked with green triangles (labeled 1–3, respectively).

Fig. 1.

Map showing the weather radar (gray star), climatic zones (Mediterranean, green; semiarid, red; arid, orange) and the four 60 km × 60 km locations that were studied (square areas, labeled A–D). The meteorological stations of Ein Hahoresh, Beer Sheva, and Sedom are marked with green triangles (labeled 1–3, respectively).

Four 60 km × 60 km locations (A–D in Fig. 1) were selected to represent climatological diversity along the EM (Mediterranean climate, A and B; semiarid climate, C and D) and geographical diversity (oversea, A; far inland, D). The selection was further based on the quality of the weather radar data, for example, ascertained by the distance from the weather radar, absence of known ground clutter, and beam blockages. The areal extent of the locations allows capturing the largest convective rain cells observed in the region (Peleg and Morin 2012; Belachsen et al. 2017) while preserving a climatic homogeneity in each location.

3. Data and methods

a. Weather radar system

Weather radars measure the reflectivity of rain droplets within atmospheric sampling volumes. Under hypotheses on the characteristics of the rain drops and on their spatial distribution, reflectivity can be converted to rain intensity. Whenever the formulated hypotheses are not fulfilled, estimation errors may occur, sometimes inducing systematic biases in the quantitative radar rainfall (Villarini and Krajewski 2010). Correction algorithms can be applied to the radar data to reduce the impact of these errors. Rainfall data used in this study were obtained from a single C-band (5.35-cm wavelength) non-Doppler instrument located in the Ben Gurion International Airport, Israel (Fig. 1). Records from this radar have been archived from October 1990 to March 2014, with small data gaps due to malfunction or regular maintenance, and were processed to reduce the impact of the abovementioned errors. With 24 hydrological years of corrected and gauge-adjusted estimates, this represents one of the longest homogeneous archives of weather radar estimates worldwide.

The radar used in this study records at a spatial resolution of 1.4° in the angular direction and 1 km radially and at a temporal resolution of 3–5 min per volume scan. The quantitative precipitation estimation procedure used in this study was developed explicitly for high-intensity events, combining physically based correction algorithms and quantitative adjustments based on comparisons with rain gauge measurements (Marra and Morin 2015). It includes 1) checking of antenna pointing; 2) filtering for ground clutter; and correcting for the effects of 3) wet radome attenuation, 4) attenuation due to the propagation of the radar beam, 5) beam blockage, and 6) vertical variations of reflectivity, together with a hail filter. A two-step bias adjustment was then applied combining 7) a yearly range-dependent and 8) an event-based mean field bias adjustment based on comparison with quality-checked data obtained from rain gauges. Maximal radar rain rates were limited to 150 mm h−1 in order to avoid contamination due to hail. Data used in this study were transformed to a Cartesian grid of 1 km × 1 km with fixed time intervals of 5 min. A fixed reflectivity to rain rate conversion ZR relationship of (Z in mm6 m−3, R in mm h−1), following the study by Morin and Gabella (2007), was used. It should be noted that the exponent of the chosen relationship is typical of convective-type rainfall, whereas the importance of the multiplier is minor, since multiplicative adjustments are subsequently applied. Further information about the adjustment procedures and the quantitative assessment of the radar archive can be found in Marra et al. (2014) and Marra and Morin (2015, 2018).

Three main sources of error could particularly affect the analyses presented in this study: 1) attenuation of the radar beam in heavy rain, 2) observation geometry, and 3) an increase of radar sampling volume with distance from the radar. All of these aspects are taken into account in the correction procedures applied to the data. Attenuation is corrected using a forward algorithm with an upper limit to the correction factor (10 dBZ) to avoid well-known numerical instabilities (Marra and Morin 2015); its effectiveness can be potentially reduced in the presence of extremely high intensities, but assessment of the data after the application of the correction (here and in other studies) suggested that this represents a problem only in a limited number of cases (see Marra et al. 2014, 2017; Marra and Morin 2015, 2018). The radar typically scans ~13 elevation angles above the horizontal plane, providing three-dimensional information on the atmospheric reflectivity. All the available vertical information is exploited to identify the vertical profile of reflectivity on 6-h-wide running windows. Assuming stationarity in the 6 h, the identified profile is used to correct the three-dimensional estimates by means of the height of the sampled volume (Marra and Morin 2015). Elevations up to 5° are then used to estimate the ground rainfall, meaning that the atmosphere is sampled at heights up to 3000 m in location D (90–140 km from the radar) and at lower heights in locations A–C. Similarly, location D experiences an increased size of the radar sampling volumes and is thus more affected by this issue. A range-dependent bias adjustment is applied to decrease these residual effects (Morin and Gabella 2007). An indirect, but critical, assessment of the quality of this radar archive for high-intensity events is provided in Marra et al. (2017), where extreme value analyses based on annual maxima from this archive and from a satellite-based precipitation product (thus completely free from attenuation and range effects) are compared, with surprisingly good spatial correlations.

b. Convective rain cells identification

Convective rain cells were identified from the radar rain fields using the image processing algorithm presented by Peleg and Morin (2012) [see Fig. 2 in Peleg and Morin (2012) for an illustrative example]. First, radar grid cells were set to a null value if the grid cells have a rain intensity lower than 10 mm h−1 or if five or more grid cells in the 3 × 3 gridcell neighborhood were null. The rain intensity threshold of 10 mm h−1, corresponding to ~40 dBZ, was set in order to capture only the convective part of the rain from the image, as was previously suggested by others for the same purpose (Rigo and Llasat 2004; Kyznarová and Novák 2009; Barnolas et al. 2010; Peleg and Morin 2012). Then, convective rain cells were identified from the rain fields as spatially connected areas greater than or equal to 9 km2. Cells with a lower area were removed in order to avoid small cells that are potentially noise (Peleg and Morin 2012). Moreover, convective rain cells not completely included in the inspected locations (i.e., extending to the boundaries) were excluded. Convective rain cells were treated independently in time (i.e., no tracking algorithm was applied), meaning that an individual rain cell is sampled multiple times as it passes through the domain. The number of identified convective rain cells varies between 6019 for location D and 86 280 for location B. Note that the identification of convective rain cells is sensitive to the applied rain intensity threshold, particularly for the dry regions; for example, the number of identified convective rain cells dropped to 5124 at location D when increasing the rain intensity threshold by 1 mm h−1 to a value of 11 mm h−1. Nevertheless, results for thresholds within 9–14 mm h−1 remain within the uncertainty range that is calculated using the adopted rain intensity threshold of 10 mm h−1.

c. Near-surface air and dew point temperatures

Near-surface air temperature (simply temperature from here on) was derived from subdaily (3, 6, or 12 h) ground stations operated by the Israel Meteorology Service (IMS). Temperature for the study locations was derived as the areal average computed every 3 h using an inverse distance weighting interpolation (as Kurtzman and Kadmon 1999) and accounting for the effect of elevation and macrogeographical variability on the lapse rate. Near-surface dewpoint temperature was derived from three IMS ground stations, located in different climate regions (Fig. 1), at 3-h intervals. The period of available temperature data overlaps with the period of the radar archive.

The analyses presented in this study are limited to the 5°–25°C temperature range in order to avoid considerable contamination of the radar estimates caused by solid precipitation (at lower temperatures) and evaporation of the rain drops during their fall from the rain cloud to the ground (at higher temperatures), the latter being particularly relevant in semiarid and arid climates. Moreover, at higher temperatures, the scaling dependence between peak rain intensity and temperature is expected to break because of the limitations in the atmospheric moisture availability, as has been shown by Drobinski et al. (2016) for the Mediterranean region and has been reported for other locations (e.g., Berg et al. 2009; Hardwick Jones et al. 2010; Panthou et al. 2014; Molnar et al. 2015). In this regard, the relationship between relative humidity and temperature for locations B and D shows humidity limitations for temperatures above 25°C (Fig. S1 in the online supplemental material), implying that the breaking point in the scaling is likely to occur also in this study area.

d. Convective available potential energy

Convective available potential energy (CAPE) was calculated using a computer script written by George Bryan (Bryan and Fritsch 2004). Air temperature and relative humidity data at 25 geopotential heights (intervals of 25 hPa for the range between 700 and 1000 hPa and intervals of 50 hPa for the range between 100 and 700 hPa) were extracted for the same period of the radar archive from the MERRA reanalysis dataset (Rienecker et al. 2011). Dewpoint temperature was calculated for each layer. The MERRA grid size is approximately 50 km × 66 km in space and 3 h in time. For each location, a time series of the spatially averaged CAPE was calculated. The relationship between the peak rainfall intensity of convective rain cells (for the 99th percentile, denoted as P99) and CAPE was also explored. A linear regression between the natural logarithmic of the peak rainfall intensity and CAPE, in the form of (as in Lepore et al. 2015, 2016), was calculated using the quantile regression method that is discussed in the next section.

e. Relationship between rainfall and temperature

The area and peak rainfall intensity of convective rain cells were scaled with temperature. The peak rainfall intensity of convective rain cells was defined as the 9 km2 spatially connected area of each cell characterized by the largest average intensity. This threshold, matching the minimum area used for the identification of convective rain cells, allows one to compare the peak intensity of rain cells on equally wide areas, limiting the possible impact of measurement errors related to weather radar estimates.

The wet area ratio, areal rainfall, convective areal rainfall, and nonconvective areal rainfall were all calculated for convective rainfall fields, that is, for radar rainfall fields in which at least one convective rain cell is observed. Wet area ratio is defined as the fraction of radar grid cells within a 60 km × 60 km domain in which rain intensity exceeds 0.1 mm h−1. Areal rainfall is defined as the average rainfall intensity of all grid cells with precipitation above 0.1 mm h−1, while convective and nonconvective areal rainfall are the average rainfall intensity of all convective rain cells and of the nonconvective cells in a convective field, respectively. This allows one to quantify the impact of temperature on the generation of total rainfall and on the relative contribution of convective and nonconvective elements.

Each convective rain cell was associated with the last temperature observation available before the rain cell detection time. The scaling of area and peak rainfall intensity of the convective rain cells (for the median and the 99th percentile) and of the wet area ratio and areal rainfall (median) with temperature was then visually analyzed using 3°C-wide bins calculated using 1°C running intervals (e.g., Berg et al. 2013; Drobinski et al. 2018) and objectively quantified using the quantile regression technique (Koenker and Bassett 1978). The latter allows us to directly estimate the scaling of a variable with temperature avoiding biases due to the definition of the bins and the number of elements sampled in each bin (Wasko and Sharma 2014; Boessenkool et al. 2017). A bootstrapping procedure was used to calculate the uncertainty related to the estimated scaling coefficients and was quantified as the 5th–95th quantile interval of 100 bootstraps with replacement from the original pool of observed data.

The two-dimensional spatial autocorrelation structure of the convective rainfall fields was calculated through the power spectral densities using the fast Fourier transform (e.g., Nerini et al. 2017). A three-parameter exponential function, , where d represents the distance, is the zero-distance correlation, is the correlation distance, and is the shape factor (e.g., Villarini et al. 2008; Peleg et al. 2013; Marra and Morin 2018; Peleg et al. 2018), was fitted to the one-dimensional correlogram. For each binned temperature interval, the correlation distance, that is, the distance at which the spatial correlation decreases by an e-folding length, was then calculated and averaged. An illustrative example of this procedure is reported in Fig. S2.

4. Results

a. Scaling of convective rain cell properties with temperature

In general, peak rain intensity consistently increases with temperature both in its extreme (99th percentile) and median values, whereas the area of convective rain cells shows a slight decrease or remains unchanged with increasing temperatures (Figs. 2a,b, Table 1).

Fig. 2.

(a) Convective cells area and (b) peak intensity as a function of near-surface air temperature. The scaling is presented for the median (q50) and for the upper 99th percentile (q99) using quantile regression (lines) and binned data for fixed temperature intervals (dots). The median values of (c) wet area ratio, (d) areal rainfall, (e) nonconvective areal rainfall, and (f) convective areal rainfall for rain fields with convective activity as a function of temperature. Empty markers are plotted for bins with a sample size of less than 500 convective cells. Different colors and symbols are for locations A–D. The shaded area in (f) presents the mean ratio of the area of convective rainfall to the stormwide area as a function of temperature averaged for the four locations.

Fig. 2.

(a) Convective cells area and (b) peak intensity as a function of near-surface air temperature. The scaling is presented for the median (q50) and for the upper 99th percentile (q99) using quantile regression (lines) and binned data for fixed temperature intervals (dots). The median values of (c) wet area ratio, (d) areal rainfall, (e) nonconvective areal rainfall, and (f) convective areal rainfall for rain fields with convective activity as a function of temperature. Empty markers are plotted for bins with a sample size of less than 500 convective cells. Different colors and symbols are for locations A–D. The shaded area in (f) presents the mean ratio of the area of convective rainfall to the stormwide area as a function of temperature averaged for the four locations.

Table 1.

Scaling values (%°C−1) of the median (q50) and the 99th percentile (q99) of area and peak rainfall intensity of the convective rain cells (Figs. 2a,b). The 5th–95th quantile uncertainty range is indicated in parentheses.

Scaling values (%°C−1) of the median (q50) and the 99th percentile (q99) of area and peak rainfall intensity of the convective rain cells (Figs. 2a,b). The 5th–95th quantile uncertainty range is indicated in parentheses.
Scaling values (%°C−1) of the median (q50) and the 99th percentile (q99) of area and peak rainfall intensity of the convective rain cells (Figs. 2a,b). The 5th–95th quantile uncertainty range is indicated in parentheses.

Small differences are observed among the locations. In locations A–C (from Mediterranean to semiarid climates and from offshore to near coastline), the intensification is more pronounced for the median values, with a slight decrease of the area (from −0.5% to −0.9%°C−1) and an increase of the peak rainfall intensity (2.7%–3.5%°C−1), whereas the extreme values show a slightly larger decrease of the area (up to −1.4%°C−1) and a smaller increase of the peak rainfall intensity (1.3%–2.4%°C−1). The results for location D (semiarid climate, far inland) are subject to the largest uncertainties (>0.8%°C−1) because of the lower number of observations, especially for higher temperatures (>20°C), and the large distance from the weather radar. Nevertheless, peak rain intensity was found to consistently increase with temperature (3.3% and 4.3%°C−1 for median and extreme rainfall, respectively). The results for the scaling of the area are not conclusive.

This positive scaling of the median and extreme peak rainfall intensity with temperature contrasts the negative scaling reported by Drobinski et al. (2018) for similar climatic regions in Israel and Greece. However, Drobinski et al. (2018) computed the scaling for 3-hourly to daily resolutions, while here a much finer temporal scale (5 min), consistent with the short life duration of the convective scales (a mean of 14 min for cells that originated from Mediterranean cyclones; Peleg and Morin 2012), is examined. The presented results indicate that the peak rainfall intensity scales at a much lower rate (1.3%–4.3%°C−1) than the ~7%°C−1 scaling expected from the Clausius–Clapeyron relation (Trenberth et al. 2003).

b. Thermodynamic and dynamic conditions

The probability distribution function of the occurrence of convective rain cells is shown in Fig. 3a. All locations share similar distributions with a peak in occurrence at 11°–13°C. Convective rain cells are less frequent at temperatures warmer than 20°C as summers in the region are dry. The observed similarity in the occurrence of convective rain cells between locations (Fig. 3a) suggests homogeneity of the thermodynamic conditions over the four locations. Averaged CAPE values were calculated for each of the locations, and their value is used as a thermodynamic indicator for the generation and sustainment of convective rain cells. The averaging was conducted only for rain fields with convective activity for which CAPE is larger than zero (~65% of the rain fields with convective activity). The average CAPE was similar between locations (Fig. 3b), with considerable overlapping of the 25th–75th percentile range (100–400 J kg−1), supporting the finding that thermodynamic conditions are relatively homogenous over the EM according to the MERRA reanalysis.

Fig. 3.

(a) The probability distribution function of convective rain cell occurrence as a function of near-surface air temperature. Different colors and symbols are for locations A–D. (b) Box-and-whisker plot showing the median (labeled), 25th–75th percentile range (shaded area), 5th–95th percentile range (bounded with lines), and outliers of CAPE values of convective rain cells for locations A–D. The analysis presented is only for convective rain cells with CAPE larger than zero; the lower number corresponds to the total number of events with a CAPE value.

Fig. 3.

(a) The probability distribution function of convective rain cell occurrence as a function of near-surface air temperature. Different colors and symbols are for locations A–D. (b) Box-and-whisker plot showing the median (labeled), 25th–75th percentile range (shaded area), 5th–95th percentile range (bounded with lines), and outliers of CAPE values of convective rain cells for locations A–D. The analysis presented is only for convective rain cells with CAPE larger than zero; the lower number corresponds to the total number of events with a CAPE value.

The relationship between the peak rainfall intensity of convective rain cells and CAPE shows similarity in the slope of the regression (α values around 0.05) between the locations (Fig. 4, Table 2). The low α values indicate that there is little dependence between the peak rainfall intensity and CAPE. We note that the α values might be exposed to artifacts, as some of the convective rain cells are generated outside of the domain and advect toward it; in other words, CAPE values cannot be attributed to the place and time of emergence of the convective cells. This might lead to an underestimation in the α values. The values of the slope (≈0.05) are particularly low, as α is expected to be in the range of 0.2–0.4 [e.g., Lepore et al. (2015), central and eastern areas of the United States] and up to 0.5 (the theoretical value); yet, this low value is not an exception, since similar low values were recently reported for the western region of the United States (Lepore et al. 2016). This finding implies that other, nonthermodynamic conditions are influencing extreme rainfall intensities. The lower intercept (β values) in the semiarid climate compared to the Mediterranean climate (of 62 versus 72 mm h−1, respectively; Table 2) suggests that dry climates are more sensitive to nonthermodynamic conditions, as discussed below.

Fig. 4.

Regression of the 99th percentile of the peak rainfall intensity of convective rain cells with CAPE for locations A–D using quantile regression. Axes are transformed using natural logarithm.

Fig. 4.

Regression of the 99th percentile of the peak rainfall intensity of convective rain cells with CAPE for locations A–D using quantile regression. Axes are transformed using natural logarithm.

Table 2.

Slope coefficients of the 99th percentile of peak rainfall intensity of the convective rain cells vs CAPE on a natural logarithm basis (Fig. 4). The 5th–95th quantile uncertainty range is indicated in parentheses.

Slope coefficients of the 99th percentile of peak rainfall intensity of the convective rain cells vs CAPE on a natural logarithm basis (Fig. 4). The 5th–95th quantile uncertainty range is indicated in parentheses.
Slope coefficients of the 99th percentile of peak rainfall intensity of the convective rain cells vs CAPE on a natural logarithm basis (Fig. 4). The 5th–95th quantile uncertainty range is indicated in parentheses.

The dynamic conditions are expected to suppress the intensification of convective rain cells with increasing temperatures due to the limitations in the availability of atmospheric moisture. An inflection point in the relationship between dewpoint temperature and air temperature would support the moisture-limited hypothesis (Hardwick Jones et al. 2010; Lenderink et al. 2011; Lepore et al. 2015) and will highlight the differences in dynamic conditions between the climatic regions. Quantile plots of dewpoint temperature against air temperature for the 25th, 50th, 75th, and 99th percentiles were plotted for three ground stations representing Mediterranean, semiarid, and arid climates in the study region (see Fig. 1 for the locations and Fig. 5 for the plots). Dynamic and humidity conditions change when moving from Mediterranean to arid climates. For the Mediterranean climate, the 99th percentile of the dewpoint temperature closely follows the 1:1 slope with air temperature, and the inflection point (detected by visual inspection) is found at about 24°C. The inflection point is lower for the semiarid climate (22.5°C) and is practically undetectable for the arid climate. Moreover, the spread of the quantiles increases when moving from the Mediterranean toward the arid region. As expected, moisture availability decreases with distance from the Mediterranean Sea, consistent with the findings of Hardwick Jones et al. (2010) and Lepore et al. (2015) that observed decreasing moisture availability with distance from the ocean.

Fig. 5.

Dewpoint temperature–air temperature quantile plots for the three meteorological stations presented in Fig. 1. The vertical line represents the inflection point from the 1:1 line for the 99th quantile, detected by visual inspection.

Fig. 5.

Dewpoint temperature–air temperature quantile plots for the three meteorological stations presented in Fig. 1. The vertical line represents the inflection point from the 1:1 line for the 99th quantile, detected by visual inspection.

c. Scaling of storm properties with temperature

The relationship between temperature and the stormwide rainfall, conditioned on the presence of convective features, is shown in Figs. 2c–f and Table S1. The wet area ratio (Fig. 2c) generally decreases with increasing temperatures while the areal rainfall (Fig. 2d) remains the same or decreases moderately. An exception is found at location D, where a clear trend cannot be identified (uncertainties cover both positive and negative trends) as a smaller number of convective events is available. The combination of a large decrease in wet area ratio with a moderate decrease in intensity of areal rainfall indicates that the total rainfall amount at the storm scale decreases considerably with increased temperature.

The nonconvective areal rainfall is found to decrease with increasing temperatures (Fig. 2e). Conversely, the convective areal rainfall increases with temperature for all locations, at least for temperatures below 20°C, and seems to stabilize at warmer temperatures (Fig. 2f). The convective areal rainfall is higher for locations A and B in comparison to locations C and D (Fig. 2f). This is expected, as the climate is changing from Mediterranean to semiarid and from near the coastline to far inland areas, where the (negative) dynamic contribution to the scaling is expected to be more pronounced. The relative areal extent of the convective rainfall on the total areal rainfall increases with temperature (Fig. 2f and Fig. S3 for all locations); since areal rainfall is rather stable, this implies that convective rainfall contributes proportionally more to the total rainfall at higher temperatures.

d. Storm spatial structure response to increasing temperatures

The changes in storm properties with increasing temperature are expected to impact the spatial structure of the rainfall fields, as the rainfall is expected to be less uniform in space. As shown using model simulations by Moseley et al. (2016) and Haerter et al. (2017), scattered cells are expected to intensify, and the remaining area of the storm is expected to reduce its extent and intensity. The observed correlation distances as a function of temperature confirm this reasoning (Fig. 6). For all the observed locations, the correlation distance decreases with temperature up to 20°C. Less clear trends can be detected for warmer temperatures, due to the limited sample size. Similar correlation distances are observed for all locations, with the exception of location A, in which, for temperatures lower than 15°C, the correlation distances are considerably shorter. This is likely related to the fact that convective cells are more intense above the sea (Fig. 2e; Karklinsky and Morin 2006; Peleg and Morin 2012).

Fig. 6.

Average correlation distance of the regional rainfall as a function of near-surface air temperature for locations A–D. Empty markers indicate bins with a sample size of less than 500 convective cells. The correlation presented is only for rain fields with convective activity.

Fig. 6.

Average correlation distance of the regional rainfall as a function of near-surface air temperature for locations A–D. Empty markers indicate bins with a sample size of less than 500 convective cells. The correlation presented is only for rain fields with convective activity.

5. Discussion

a. Enhanced moisture convergence at warmer temperatures

The reduced stormwide wet area ratio and the intensification of convective rain cells at the expense of lower-intensity rainfall at warmer temperatures implies a redistribution of part of the moisture that is potentially available in the atmospheric column from the low-intensity rainfall zones toward the convective rain cells. The enhanced convective mechanism contributes to the reduction of the total rainfall amount that is observed at warmer temperatures. Observations of moisture redistribution from the storm boundaries to the center were also reported by Wasko et al. (2016b) at the storm scale and by Trenberth et al. (2003) at larger, regional scales. Here, we bring observation-based evidence of a redistribution of the moisture at much smaller scales (i.e., the scales of scattered convective rain cells: 101–103 km2), which is consistent with the model-based results reported by Loriaux et al. (2013), where simulations using a plume model show an enhancement of convective convergence at warmer temperatures.

Further investigation is needed to elucidate the relative contribution of the thermodynamic and dynamic processes and to analyze the effect of other aspects of the storm environment on the convective moisture convergence. However, this type of analysis will require the use of convection-permitting climate models (Prein et al. 2015, 2017b,a) as gridded climate variables are in general not available at the fine scales typical of the convective structures.

b. Geographic and climatic differences

The relationship between convective rainfall and temperature was explored at the minute scale (i.e., up to 5 min) for different climates and regions. The scaling for extreme convective rainfall (often defined as the slope of the 99th percentile of rainfall intensity) was mostly reported to be in the range between the CC relation (7%°C−1) and 2 × CC (14%°C−1), for example, in Austria (Schroeer and Kirchengast 2018), Germany (Berg et al. 2013; Moseley et al. 2013), Spain (Drobinski et al. 2018), Switzerland (Molnar et al. 2015), and the Netherlands (Lochbihler et al. 2017). The presented results indicate that the rates can be much lower, but still positive, when considering Mediterranean (1.3%–1.9%°C−1) and semiarid (2.4%–4.3%°C−1) climates. Spatial differences in the scaling are expected because of differences in the dynamic contribution; see, for example, the global maps of scaling by Wasko et al. (2016a) for a temporal resolution of 3 h and the role of the dynamic contribution in suppressing positive scaling in Mediterranean regions (Pfahl et al. 2017). As the climate and thermodynamic/dynamic conditions are similar between the EM regions and wide areas in North Africa (e.g., Egypt and Libya), south of Italy, in Greece, and west of Turkey, it is likely that the scaling trends reported for the Mediterranean and semiarid climate here will be valid there as well.

The enhanced convective moisture convergence observed at warmer temperatures for the Mediterranean and semiarid regions was also observed for temperate and (to a lesser extent) arid climates in Australia (Wasko et al. 2016b) and over Germany (Berg et al. 2013). However, Lochbihler et al. (2017) did not find any evidence of a redistribution of moisture toward convective rain cells in the Netherlands. The negative contribution of the dynamic component in the analyzed region and certain parts of Australia can potentially reconcile conflicting results between these studies. Further investigation is needed to understand if other aspects of the storm environment (e.g., vertical pressure velocity) could affect the size of the storm/convective cells with increasing temperatures.

c. Implications on future climate

Results presented in this study are based on current climate observations. In a warmer future climate, less rainfall is expected on average for the eastern Mediterranean region at the storm scale (e.g., Peleg et al. 2015; Samuels et al. 2018) because of a poleward propagation of the storm tracks (e.g., Tamarin-Brodsky and Kaspi 2017), while the changes to the characteristics of convective rain cells are substantially unknown and are yet to be explored. Will individual convective rain cells become more dominant and intense in a future climate? Interpretation of the results from this study may imply so, but, as the dynamic of the atmosphere will be likely different, there is no guarantee that a scaling with present-day temperatures will continue to hold true in the future. Pfahl et al. (2017), for example, used climate models to show that thermodynamic conditions will lead to a regional spatially homogeneous increase in extreme precipitation (~9%°C−1, for daily extreme rainfall) over the EM, while the dynamic contribution in this area will act in an opposite direction (~−6%°C−1) and thus will weaken the scaling of extreme rainfall. Moreover, changes in convective rainfall intensity and occurrence due to changes in climate dynamics are already observed in some parts of the world (e.g., Feng et al. 2016; Ye et al. 2017). High-resolution, convection-resolving climate models can be used to supply accurate information on the projections of changes in convective rainfall, as these types of models explicitly simulate the future dynamics of the atmosphere (both thermodynamic and dynamic conditions), allowing us to evaluate the possible uncertainties and biases related to the extrapolation of the present-day scaling features to future climates (e.g., Ban et al. 2015; Prein et al. 2017a,b).

6. Conclusions

Examination of a long-term weather radar archive in the eastern Mediterranean showed evidence of an intensification of convective rain cells with warmer near-surface air temperature at the minute scale. Extreme and median peak rainfall intensities of the convective rain cells increase with temperature, while the area slightly decreases or, at most, remains unchanged. The convective areal rainfall was found to increase with warmer temperatures while the areal nonconvective rainfall and the stormwide area decrease. These observations likely point to a redistribution of the available atmospheric moisture from the entire storm toward the convective rain cells due to an enhanced convective activity, resulting in a reduction of total rainfall amount in the entire area at warmer temperatures and a reduced spatial correlation distance of the rainfall fields. Thermodynamic conditions generating convective rain cells were found to be relatively homogenous within the region, whereas dynamic conditions were observed to suppress convection when moving from the main source of moisture, the Mediterranean Sea, toward inland drier areas, that is, reduced convective intensity occurs with reduced moisture availability.

Acknowledgments

The study was funded by the Swiss National Science Foundation (International Short Visits, project IZK0Z2_173679), by the Lady Davis Fellowship Trust (project: RainFreq) and by the Israel Science Foundation (Grant 1007/15). The authors thank E.M.S. (Mekorot Company) for the radar data and the Israel Meteorology Service for the temperature data. The authors gratefully acknowledge the useful discussions with Maya Bartov.

REFERENCES

REFERENCES
Ban
,
N.
,
J.
Schmidli
, and
C.
Schaer
,
2015
:
Heavy precipitation in a changing climate: Does short-term summer precipitation increase faster?
Geophys. Res. Lett.
,
42
,
1165
1172
, https://doi.org/10.1002/2014GL062588.
Bao
,
J.
,
S. C.
Sherwood
,
L. V.
Alexander
, and
J. P.
Evans
,
2017
:
Future increases in extreme precipitation exceed observed scaling rates
.
Nat. Climate Change
,
7
,
128
132
, https://doi.org/10.1038/nclimate3201.
Barnolas
,
M.
,
T.
Rigo
, and
M. C.
Llasat
,
2010
:
Characteristics of 2-D convective structures in Catalonia (NE Spain): An analysis using radar data and GIS
.
Hydrol. Earth Syst. Sci.
,
14
,
129
139
, https://doi.org/10.5194/hess-14-129-2010.
Belachsen
,
I.
,
F.
Marra
,
N.
Peleg
, and
E.
Morin
,
2017
:
Convective rainfall in a dry climate: Relations with synoptic systems and flash-flood generation in the Dead Sea region
.
Hydrol. Earth Syst. Sci.
,
21
,
5165
5180
, https://doi.org/10.5194/hess-21-5165-2017.
Berg
,
P.
,
J. O.
Haerter
,
P.
Thejll
,
C.
Piani
,
S.
Hagemann
, and
J. H.
Christensen
,
2009
:
Seasonal characteristics of the relationship between daily precipitation intensity and surface temperature
.
J. Geophys. Res.
,
114
,
D18102
, https://doi.org/10.1029/2009JD012008.
Berg
,
P.
,
C.
Moseley
, and
J. O.
Haerter
,
2013
:
Strong increase in convective precipitation in response to higher temperatures
.
Nat. Geosci.
,
6
,
181
185
, https://doi.org/10.1038/ngeo1731.
Boessenkool
,
B.
,
G.
Burger
, and
M.
Heistermann
,
2017
:
Effects of sample size on estimation of rainfall extremes at high temperatures
.
Nat. Hazards Earth Syst. Sci.
,
17
,
1623
1629
, https://doi.org/10.5194/nhess-17-1623-2017.
Bryan
,
G. H.
, and
J. M.
Fritsch
,
2004
:
A reevaluation of ice–liquid water potential temperature
.
Mon. Wea. Rev.
,
132
,
2421
2431
, https://doi.org/10.1175/1520-0493(2004)132<2421:AROIWP>2.0.CO;2.
Drobinski
,
P.
,
B.
Alonzo
,
S.
Bastin
,
N. D.
Silva
, and
C.
Muller
,
2016
:
Scaling of precipitation extremes with temperature in the French Mediterranean region: What explains the hook shape?
J. Geophys. Res. Atmos.
,
121
,
3100
3119
, https://doi.org/10.1002/2015JD023497.
Drobinski
,
P.
, and Coauthors
,
2018
:
Scaling precipitation extremes with temperature in the Mediterranean: Past climate assessment and projection in anthropogenic scenarios
.
Climate Dyn.
, https://doi.org/10.1007/S00382-016-3083-X,
in press
.
Eltahir
,
E. A. B.
, and
J. S.
Pal
,
1996
:
Relationship between surface conditions and subsequent rainfall in convective storms
.
J. Geophys. Res.
,
101
,
26 237
26 245
, https://doi.org/10.1029/96JD01380.
Fatichi
,
S.
,
P.
Molnar
,
T.
Mastrotheodoros
, and
P.
Burlando
,
2015
:
Diurnal and seasonal changes in near-surface humidity in a complex orography
.
J. Geophys. Res. Atmos.
,
120
,
2358
2374
, https://doi.org/10.1002/2014JD022537.
Feng
,
Z.
,
L. R.
Leung
,
S.
Hagos
,
R. A.
Houze
,
C. D.
Burleyson
, and
K.
Balaguru
,
2016
:
More frequent intense and long-lived storms dominate the springtime trend in central US rainfall
.
Nat. Commun.
,
7
,
13429
, https://doi.org/10.1038/ncomms13429.
Fischer
,
E. M.
, and
R.
Knutti
,
2016
:
Observed heavy precipitation increase confirms theory and early models
.
Nat. Climate Change
,
6
,
986
991
, https://doi.org/10.1038/nclimate3110.
Goldreich
,
Y.
,
2012
: The Climate of Israel: Observation, Research and Application. Springer, 270 pp.
Haerter
,
J. O.
,
P.
Berg
, and
C.
Moseley
,
2017
:
Precipitation onset as the temporal reference in convective self-organization
.
Geophys. Res. Lett.
,
44
,
6450
6459
, https://doi.org/10.1002/2017GL073342.
Hardwick Jones
,
R.
,
S.
Westra
, and
A.
Sharma
,
2010
:
Observed relationships between extreme sub-daily precipitation, surface temperature, and relative humidity
.
Geophys. Res. Lett.
,
37
,
L22805
, https://doi.org/10.1029/2010GL045081.
Kahana
,
R.
,
B.
Ziv
,
Y.
Enzel
, and
U.
Dayan
,
2002
:
Synoptic climatology of major floods in the Negev Desert, Israel
.
Int. J. Climatol.
,
22
,
867
882
, https://doi.org/10.1002/joc.766.
Karklinsky
,
M.
, and
E.
Morin
,
2006
:
Spatial characteristics of radar-derived convective rain cells over southern Israel
.
Meteor. Z.
,
15
,
513
520
, https://doi.org/10.1127/0941-2948/2006/0153.
Koenker
,
R.
, and
G.
Bassett
,
1978
:
Regression quantiles
.
Econometrica
,
46
,
33
50
, https://doi.org/10.2307/1913643.
Kurtzman
,
D.
, and
R.
Kadmon
,
1999
:
Mapping of temperature variables in Israel: A comparison of different interpolation methods
.
Climate Res.
,
13
,
33
43
, https://doi.org/10.3354/cr013033.
Kyznarová
,
H.
, and
P.
Novák
,
2009
:
CELLTRACK—Convective cell tracking algorithm and its use for deriving life cycle characteristics
.
Atmos. Res.
,
93
,
317
327
, https://doi.org/10.1016/j.atmosres.2008.09.019.
Lenderink
,
G.
,
H. Y.
Mok
,
T. C.
Lee
, and
G. J.
van Oldenborgh
,
2011
:
Scaling and trends of hourly precipitation extremes in two different climate zones - Hong Kong and the Netherlands
.
Hydrol. Earth Syst. Sci.
,
15
,
3033
3041
, https://doi.org/10.5194/hess-15-3033-2011.
Lepore
,
C.
,
D.
Veneziano
, and
A.
Molini
,
2015
:
Temperature and CAPE dependence of rainfall extremes in the eastern United States
.
Geophys. Res. Lett.
,
42
,
74
83
, https://doi.org/10.1002/2014GL062247.
Lepore
,
C.
,
J. T.
Allen
, and
M. K.
Tippett
,
2016
:
Relationships between hourly rainfall intensity and atmospheric variables over the contiguous United States
.
J. Climate
,
29
,
3181
3197
, https://doi.org/10.1175/JCLI-D-15-0331.1.
Lochbihler
,
K.
,
G.
Lenderink
, and
A. P.
Siebesma
,
2017
:
The spatial extent of rainfall events and its relation to precipitation scaling
.
Geophys. Res. Lett.
,
44
,
8629
8636
, https://doi.org/10.1002/2017GL074857.
Loriaux
,
J. M.
,
G.
Lenderink
,
S. R.
De Roode
, and
A. P.
Siebesma
,
2013
:
Understanding convective extreme precipitation scaling using observations and an entraining plume model
.
J. Atmos. Sci.
,
70
,
3641
3655
, https://doi.org/10.1175/JAS-D-12-0317.1.
Marra
,
F.
, and
E.
Morin
,
2015
:
Use of radar QPE for the derivation of intensity–duration–frequency curves in a range of climatic regimes
.
J. Hydrol.
,
531
,
427
440
, https://doi.org/10.1016/j.jhydrol.2015.08.064.
Marra
,
F.
, and
E.
Morin
,
2018
:
Autocorrelation structure of convective rainfall in semiarid-arid climate derived from high-resolution X-band radar estimates
.
Atmos. Res.
,
200
,
126
138
, https://doi.org/10.1016/j.atmosres.2017.09.020.
Marra
,
F.
,
E. I.
Nikolopoulos
,
J. D.
Creutin
, and
M.
Borga
,
2014
:
Radar rainfall estimation for the identification of debris-flow occurrence thresholds
.
J. Hydrol.
,
519
,
1607
1619
, https://doi.org/10.1016/j.jhydrol.2014.09.039.
Marra
,
F.
,
E.
Morin
,
N.
Peleg
,
Y.
Mei
, and
E. N.
Anagnostou
,
2017
:
Intensity-duration-frequency curves from remote sensing rainfall estimates: Comparing satellite and weather radar over the eastern Mediterranean
.
Hydrol. Earth Syst. Sci.
,
21
,
2389
2404
, https://doi.org/10.5194/hess-21-2389-2017.
Meredith
,
E. P.
,
V. A.
Semenov
,
D.
Maraun
,
W.
Park
, and
A. V.
Chernokulsky
,
2015
:
Crucial role of Black Sea warming in amplifying the 2012 Krymsk precipitation extreme
.
Nat. Geosci.
,
8
,
615
619
, https://doi.org/10.1038/ngeo2483.
Mishra
,
V.
,
J. M.
Wallace
, and
D. P.
Lettenmaier
,
2012
:
Relationship between hourly extreme precipitation and local air temperature in the United States
.
Geophys. Res. Lett.
,
39
,
L16403
, https://doi.org/10.1029/2012GL052790.
Molnar
,
P.
,
S.
Fatichi
,
L.
Gaal
,
J.
Szolgay
, and
P.
Burlando
,
2015
:
Storm type effects on super Clausius-Clapeyron scaling of intense rainstorm properties with air temperature
.
Hydrol. Earth Syst. Sci.
,
19
,
1753
1766
, https://doi.org/10.5194/hess-19-1753-2015.
Morin
,
E.
, and
M.
Gabella
,
2007
:
Radar-based quantitative precipitation estimation over Mediterranean and dry climate regimes
.
J. Geophys. Res.
,
112
,
D20108
, https://doi.org/10.1029/2006JD008206.
Morin
,
E.
,
D.
Goodrich
,
R.
Maddox
,
X.
Gao
,
H.
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.
Morin
,
E.
,
Y.
Jacoby
,
S.
Navon
, and
E.
Bet-Halachmi
,
2009
:
Towards flash-flood prediction in the dry Dead Sea region utilizing radar rainfall information
.
Adv. Water Resour.
,
32
,
1066
1076
, https://doi.org/10.1016/j.advwatres.2008.11.011.
Moseley
,
C.
,
P.
Berg
, and
J. O.
Haerter
,
2013
:
Probing the precipitation life cycle by iterative rain cell tracking
.
J. Geophys. Res. Atmos.
,
118
,
13 361
13 370
, https://doi.org/10.1002/2013JD020868.
Moseley
,
C.
,
C.
Hohenegger
,
P.
Berg
, and
J. O.
Haerter
,
2016
:
Intensification of convective extremes driven by cloud-cloud interaction
.
Nat. Geosci.
, 9,
748
752
, https://doi.org/10.1038/ngeo2789.
Nerini
,
D.
,
N.
Besic
,
I.
Sideris
,
U.
Germann
, and
L.
Foresti
,
2017
:
A non-stationary stochastic ensemble generator for radar rainfall fields based on the short-space Fourier transform
.
Hydrol. Earth Syst. Sci.
,
21
,
2777
2797
, https://doi.org/10.5194/hess-21-2777-2017.
O’Gorman
,
P. A.
, and
T.
Schneider
,
2009
:
Scaling of precipitation extremes over a wide range of climates simulated with an idealized GCM
.
J. Climate
,
22
,
5676
5685
, https://doi.org/10.1175/2009JCLI2701.1.
Panthou
,
G.
,
A.
Mailhot
,
E.
Laurence
, and
G.
Talbot
,
2014
:
Relationship between surface temperature and extreme rainfalls: A multi-time-scale and event-based analysis
.
J. Hydrometeor.
,
15
,
1999
2011
, https://doi.org/10.1175/JHM-D-14-0020.1.
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.
Peleg
,
N.
,
M.
Ben-Asher
, and
E.
Morin
,
2013
:
Radar subpixel-scale rainfall variability and uncertainty: Lessons learned from observations of a dense rain-gauge network
.
Hydrol. Earth Syst. Sci.
,
17
,
2195
2208
, https://doi.org/10.5194/hess-17-2195-2013.
Peleg
,
N.
,
M.
Bartov
, and
E.
Morin
,
2015
:
CMIP5-predicted climate shifts over the East Mediterranean: Implications for the transition region between Mediterranean and semi-arid climates
.
Int. J. Climatol.
,
35
,
2144
2153
, https://doi.org/10.1002/joc.4114.
Peleg
,
N.
,
F.
Marra
,
S.
Fatichi
,
A.
Paschalis
,
P.
Molnar
, and
P.
Burlando
,
2018
:
Spatial variability of extreme rainfall at radar subpixel scale
.
J. Hydrol.
,
556
,
922
933
, https://doi.org/10.1016/j.jhydrol.2016.05.033.
Pfahl
,
S.
,
P. O.
Gorman
, and
E.
Fischer
,
2017
:
Understanding the regional pattern of projected future changes in extreme precipitation
.
Nat. Climate Change
,
7
,
423
427
, https://doi.org/10.1038/nclimate3287.
Prein
,
A. F.
, and Coauthors
,
2015
:
A review on regional convection-permitting climate modeling: Demonstrations, prospects, and challenges
.
Rev. Geophys.
,
53
,
323
361
, https://doi.org/10.1002/2014RG000475.
Prein
,
A. F.
,
C.
Liu
,
K.
Ikeda
,
S. B.
Trier
,
R. M.
Rasmussen
,
G. J.
Holland
, and
M. P.
Clark
,
2017a
:
Increased rainfall volume from future convective storms in the US
.
Nat. Climate Change
,
7
,
880
884
, https://doi.org/10.1038/s41558-017-0007-7.
Prein
,
A. F.
,
R. M.
Rasmussen
,
K.
Ikeda
,
C.
Liu
,
M. P.
Clark
, and
G. J.
Holland
,
2017b
:
The future intensification of hourly precipitation extremes
.
Nat. Climate Change
,
7
,
48
52
, https://doi.org/10.1038/nclimate3168.
Rienecker
,
M. M.
, and Coauthors
,
2011
:
MERRA: NASA's Modern-Era Retrospective Analysis for Research and Applications
.
J. Climate
,
24
,
3624
3648
, https://doi.org/10.1175/JCLI-D-11-00015.1.
Rigo
,
T.
, and
M. C.
Llasat
,
2004
:
A methodology for the classification of convective structures using meteorological radar: Application to heavy rainfall events on the Mediterranean coast of the Iberian Peninsula
.
Nat. Hazards Earth Syst. Sci.
,
4
,
59
68
, https://doi.org/10.5194/nhess-4-59-2004.
Samuels
,
R.
, and Coauthors
,
2018
:
Evaluation and projection of extreme precipitation indices in the Eastern Mediterranean based on CMIP5 multi-model ensemble
.
Int. J. Climatol.
, https://doi.org/10.1002/joc.5334,
in press
.
Schroeer
,
K.
, and
G.
Kirchengast
,
2018
:
Sensitivity of extreme precipitation to temperature: The variability of scaling factors from a regional to local perspective
.
Climate Dyn.
, https://doi.org/10.1007/s00382-017-3857-9,
in press
.
Sharon
,
D.
, and
H.
Kutiel
,
1986
:
The distribution of rainfall intensity in Israel, its regional and seasonal variations and its climatological evaluation
.
J. Climatol.
,
6
,
277
291
, https://doi.org/10.1002/joc.3370060304.
Singleton
,
A.
, and
R.
Toumi
,
2013
:
Super-Clausius–Clapeyron scaling of rainfall in a model squall line
.
Quart. J. Roy. Meteor. Soc.
,
139
,
334
339
, https://doi.org/10.1002/qj.1919.
Syed
,
K. H.
,
D. C.
Goodrich
,
D. E.
Myers
, and
S.
Sorooshian
,
2003
:
Spatial characteristics of thunderstorm rainfall fields and their relation to runoff
.
J. Hydrol.
,
271
,
1
21
, https://doi.org/10.1016/S0022-1694(02)00311-6.
Tamarin-Brodsky
,
T.
, and
Y.
Kaspi
,
2017
:
Enhanced poleward propagation of storms under climate change
.
Nat. Geosci.
,
10
,
908
913
, https://doi.org/10.1038/s41561-017-0001-8.
Trenberth
,
K.
,
A.
Dai
,
R.
Rasmussen
, and
D.
Parsons
,
2003
:
The changing character of precipitation
.
Bull. Amer. Meteor. Soc.
,
84
,
1205
1217
, https://doi.org/10.1175/BAMS-84-9-1205.
Villarini
,
G.
, and
W. F.
Krajewski
,
2010
:
Review of the different sources of uncertainty in single polarization radar-based estimates of rainfall
.
Surv. Geophys.
,
31
,
107
129
, https://doi.org/10.1007/s10712-009-9079-x.
Villarini
,
G.
,
P. V.
Mandapaka
,
W. F.
Krajewski
, and
R. J.
Moore
,
2008
:
Rainfall and sampling uncertainties: A rain gauge perspective
.
J. Geophys. Res.
,
113
,
D11102
, https://doi.org/10.1029/2007JD009214.
Wasko
,
C.
, and
A.
Sharma
,
2014
:
Quantile regression for investigating scaling of extreme precipitation with temperature
.
Water Resour. Res.
,
50
,
3608
3614
, https://doi.org/10.1002/2013WR015194.
Wasko
,
C.
,
A.
Sharma
, and
F.
Johnson
,
2015
:
Does storm duration modulate the extreme precipitation-temperature scaling relationship?
Geophys. Res. Lett.
,
42
,
8783
8790
, https://doi.org/10.1002/2015GL066274.
Wasko
,
C.
,
R. M.
Parinussa
, and
A.
Sharma
,
2016a
:
A quasi-global assessment of changes in remotely sensed rainfall extremes with temperature
.
Geophys. Res. Lett.
,
43
,
12 659
12 668
, https://doi.org/10.1002/2016GL071354.
Wasko
,
C.
,
A.
Sharma
, and
S.
Westra
,
2016b
:
Reduced spatial extent of extreme storms at higher temperatures
.
Geophys. Res. Lett.
,
43
,
4026
4032
, https://doi.org/10.1002/2016GL068509.
Westra
,
S.
, and Coauthors
,
2014
:
Future changes to the intensity and frequency of short-duration extreme rainfall
.
Rev. Geophys.
,
52
,
522
555
, https://doi.org/10.1002/2014RG000464.
Yakir
,
H.
, and
E.
Morin
,
2011
:
Hydrologic response of a semi-arid watershed to spatial and temporal characteristics of convective rain cells
.
Hydrol. Earth Syst. Sci.
,
15
,
393
404
, https://doi.org/10.5194/hess-15-393-2011.
Ye
,
H.
,
E. J.
Fetzer
,
S.
Wong
, and
B. H.
Lambrigtsen
,
2017
:
Rapid decadal convective precipitation increase over Eurasia during the last three decades of the 20th century
.
Sci. Adv.
,
3
, https://doi.org/10.1126/sciadv.1600944.

Footnotes

Supplemental information related to this paper is available at the Journals Online website: https://doi.org/10.1175/JHM-D-17-0158.s1.

© 2018 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Supplemental Material