Severe flash floods in parts of Israel

By Oren Davidoff, Israel Meteorological Service (IMS) and Vesa Nietosvaara (EUMETSAT)

Between 25 and 27 April, massive storm cells hit many regions in the southern part of Israel. One of the violent cells that developed on 26 April, as seen on the Meteosat-9 Natural Colour RGB (Figure 1), caused several extreme flash flood events in the Judean Desert, the Arava and the Eastern Negev, all located in the southern part of Israel. One of the most intense flash floods on that day, in the eastern Negev, killed 10 teenagers hiking in the area.

The convective systems on the 26 April can also be seen in the Meteosat-9 HRV (Figures 2 and 3), thermal infrared (Figure 4 and 5) and Airmass RGB (Figures 6 and 7) imagery from 09:00 and 12:00 UTC. On the Airmass imagery red hues represent descending stratospheric air followed by a cold airmass, indicating a low tropopause region.

The floods occurred in the southern part of Israel (see map ), along the boundary of two arid geoclimatic zones known as the northeastern Negev and the northern Arava, where the mean annual rainfall is between 21–100 mm.

While the month of April in Israel is considered to be a transitional season, just a step away from the hot and mostly humid summer, occasional rain systems and warm lows sometimes pass over the country, causing widespread haze or dust events.

April 2018 had slightly higher temperatures than normal and higher than average precipitation in most of the country's regions, except the northeastern parts.

The precipitation fell mostly in three major events, with the most significant one occurring between 25 and 27 April.

These floods broke rainfall records and resulted in 14 deaths and widespread damage. Extensive widespread damage was inflicted on agricultural lands in the southern part of the Galilee Sea and the Beit She'an Valley, while extensive floods occurred in several major cities, including Jerusalem and Tel Aviv.

Some extreme, unprecedented rainfall values were recorded during this event:

1. 23.6 mm rainfall during 10 minutes in Hefetz Haim (translating into a fall rate of 141.6 mm/hr.)
2. Eden Farm had 19.7 mm (118.2 mm/hr.)
3. Jerusalem had 17.1 mm (102.6 mm/hr.)

The genesis of the flooding originated in a low-pressure system which situated itself over the Sinai Peninsula and southern Israel on 25 April. It peaked on 26 April between 9:00 and 12:00 UTC, see ECMWF NWP synoptic analysis layered over MODIS True Color RGB image, 26 April 12:00 UTC (Figure 8),

It slowly moved eastward, centering over Jordan and the NW Arabian Peninsula on 26 April, finally reaching eastern Jordan and western Iraq on 27 April.

The development of the low observed in the 500 hPa pressure level can be seen in the animated gif of ECMWF analysis, 25 April 06:00 UTC–26 April 12:00 UTC (Figure 9). It was well forecast by the 25 April 00:00 UTC ECMWF model run. 

The low extended into deepening troughs in the mid and low levels of the atmosphere, resulting in significant instability and other convective parameters (these were 'flagged' in the forecasts, but not necessarily pronounced enough to exactly pinpoint the area where the flash floods occurred), contributing to the development of these substantial convective cells (i.e. CAPE, shear, etc.).

The forecast lapse rate depicted by the temperature gradient between the 850 hPa and the 500 hPa pressure levels , was justification (along with other forecast parameters such as mid-level humidity) for forecasters to issue severe flash floods warnings a few days before the event.

A combination of cold air aloft, with hot and humid air at ground level, caused a significant unstable atmosphere, initiating substantial convective cells over the region. Isolated convective cells entered northern Israel on the morning of the 25th, later developing in the south.

On the morning of 26 April, convective cells from southward moving cells, reached northern Israel. During late morning severe convective cells developed over the south, in the region of the aforementioned tragedy, consistently bombarding the watershed zone of the Tsafit Stream (Figure 10).

The sequence shown in Figure 10 is a zoomed-in area of the enlarged IMS radar images in Figure 11, showing the passing of convective cells accompanied by thunderstorms and heavy rain.

In Figure 10, compared to Figure 11, the colours indicate the intensity of rain in the area, represented by the returned signal from raindrops (reflectivity) picked by our radar, in DBZ, and not the amount of rain accumulated, as the scale on Figure 11 shows.

On Figure 11 it can be seen that the majority of precipitation was concentrated in the central-eastern part of Israel, in the central mountain regions of Samaria and Judea, thus setting the grounds for the severe flash floods that occurred east of that region, in the lower parts of the Dead Sea and the Judean Desert.

The Skew-T plotted that day from the 12:00 UTC Beit Dagan radiosonde (Figure 12), supports the severity of the instability observed/measured. While the temperature gradient between the 850 hPa pressure level and the 500 hPa is needed in this region for getting well-developed convective cells (∆T~28°C), alongside other relevant parameters such as the CAPE and CIN values (*CAPE=909.1, CINS=-0.29;) the environmental humidity values seem to 'stutter' in higher elevations, leaving the forecaster with the uncertainty of whether there will be cell development, like the one that developed in the south.

*In this case, the CAPE values in Israel may not be comparable to worldwide observed values in thunderstorm environments that often may exceed 1,000 joules per kilogram (j/kg), but in this region the value is sufficient for storm initiation, while the CIN value goes with a strong development..

But, taking into consideration the location of the sonde's release (the centre of Israel), the sketched Skew-T may have only partially pre-supposed the actual severe and unstable conditions taking place in the south, where the flood occurred. The forecast tephigram plot from an area called Hazeva (Figure 13), just 35 km south of the flooded area, depicts a more realistic tally of the atmospheric conditions that were, in effect, responsible for the cell causing the flood.

The sets of four different satellite images from consecutive hours (Figures 14-17. Imagery courtesy of Elyakom Vadislavsky.) depict a precise evolution of the development of the most critical cells in the south, beginning at 09:00 UTC and ending at 12:00 UTC. In the bottom right of each of the images is the High Resolution Visible channel, bottom left is the Convective Storm RGB, top right is the Day Microphysics RGB Composite, and top left is the IR 10.8 µm, layered with a coloured scheme to depict the height of the top of the developed clouds.

Related content

Incredible Flooding at Zin River in Negev Desert, Israel (Breaking Disaster/YouTube)
Ten youth dead in Southern Israel flash floods (The Jerusalem Post)
Parents of flash flood victims demand independent inquiry (ynet News)