Small scale rainfall variability is a key factor driving runoff response in fast responding systems, such as mountainous, urban and arid catchments. In this paper, the spatial–temporal autocorrelation structure of convective rainfall is derived with extremely high resolutions (60 m, 1 min) using estimates from an X-Band weather radar recently installed in a semiarid-arid area. The 2-dimensional spatial autocorrelation of convective rainfall fields and the temporal autocorrelation of point-wise and distributed rainfall fields are examined. The autocorrelation structures are characterized by spatial anisotropy, correlation distances \~ 1.5–2.8 km and rarely exceeding 5 km, and time-correlation distances \~ 1.8–6.4 min and rarely exceeding 10 min. The observed spatial variability is expected to negatively affect estimates from rain gauges and microwave links rather than satellite and C-/S-Band radars; conversely, the temporal variability is expected to negatively affect remote sensing estimates rather than rain gauges. The presented results provide quantitative information for stochastic weather generators, cloud-resolving models, dryland hydrologic and agricultural models, and multi-sensor merging techniques.
Space-time patterns of rainfall are important climatic characteristics that influence runoff generation and flash flood magnitude. Their derivation requires high-resolution measurements to adequately represent the rainfall distribution, and is best provided by remote sensing tools. This need is further emphasized in dry climate regions, where rainfall is scarce and, often, local and highly variable. Our research is focused on understanding the nature of rainfall events in the dry Dead Sea region (Eastern Mediterranean) by identifying and characterizing the spatial structure and the dynamics of convective storm cores (known as rain cells). To do so, we take advantage of 25 years of corrected and gauge-adjusted weather radar data. A statistical analysis of convective rain-cells spatial and temporal characteristics was performed with respect to synoptic pattern, geographical location, and flash flood generation. Rain cells were extracted from radar data using a cell segmentation method and a tracking algorithm and were divided into rain events. A total of 10,500 rain cells, 2650 cell tracks and 424 rain events were elicited. Rain cell properties, such as mean areal and maximal rain intensity, area, life span, direction and speed, were derived. Rain events were clustered, according to several ERA-Interim atmospheric parameters, and associated with three main synoptic patterns: Cyprus Low, Low to the East of the study region and Active Red Sea Trough. The first two originate from the Mediterranean Sea, while the third is an extension of the African monsoon. On average, the convective rain cells in the region are 90 km 2 in size, moving from West to East in 13 ms −1 and living 18 minutes. Several significant differences between rain cells of the various synoptic types were observed. In particular, Active Red Sea Trough rain cells are characterized by higher rain intensities and lower speeds, suggesting a higher flooding potential for small catchments. The north-south negative gradient of mean annual rainfall in the study region was found to be negatively correlated with rain cells intensity and positively correlated with rain cells area. Additional analysis was done for convective rain cells over two nearby catchments located in the central part of the study region, by ascribing some of the rain events to observed flash-flood events. It was found that rain events associated with flash-floods have higher maximal rain cell intensity and lower minimal cell speed than rain events that did not lead to a flash-flood in the watersheds. This information contributes to our understanding of rain patterns over the dry area of the Dead Sea and their connection to flash-floods. The statistical distributions of rain cells properties can be used for high space-time resolution stochastic simulations of rain storms that can serve as an input to hydrological models.
The hydroclimatology, hydrometeorology, and hydrology of flash floods in the arid/semiarid southwestern United States are examined through empirical analyses of long-term, high-resolution rainfall and stream gauging observations, together with hydrological modeling analyses of the 19 August 2014 storm based on the Kinematic Runoff and Erosion Model (KINEROS2). The analyses presented here are centered on identifying the structure and evolution of flood-producing storms, as well as the interactions of space–time rainfall variability and basin characteristics in determining the upper-tail properties of rainfall and flood magnitudes over this region. This study focuses on four watersheds in Maricopa County, Arizona, with contrasting geomorphological properties. Flash floods over central Arizona are concentrated in both time and space, reflecting controls of the North American monsoon and complex terrain. Thunderstorm systems during the North American monsoon, as represented by the 19 August 2014 storm, are the dominant flood agents that determine the upper tail of flood frequency over central Arizona and that also shape the envelope curve of floods for watersheds smaller than 250 km2. Flood response for the 19 August 2014 storm is associated with storm elements of comparable spatial extent to the drainage area and slow movement for the three compact, headwater watersheds. Flood response for the elongated and relatively flat Skunk Creek highlights the importance of the spatial distribution of rainfall for transmission losses in arid/semiarid watersheds.
Intensity–duration–frequency (IDF) curves are widely used to quantify the probability of occurrence of rainfall extremes. The usual rain gauge-based approach provides accurate curves for a specific location, but uncertainties arise when ungauged regions are examined or catchment-scale information is required. Remote sensing rainfall records, e.g. from weather radars and satellites, are recently becoming available, providing high-resolution estimates at regional or even global scales; their uncertainty and implications on water resources applications urge to be investigated. This study compares IDF curves from radar and satellite (CMORPH) estimates over the eastern Mediterranean (covering Mediterranean, semiarid, and arid climates) and quantifies the uncertainty related to their limited record on varying climates. We show that radar identifies thicker-tailed distributions than satellite, in particular for short durations, and that the tail of the distributions depends on the spatial and temporal aggregation scales. The spatial correlation between radar IDF and satellite IDF is as high as 0.7 for 2–5-year return period and decreases with longer return periods, especially for short durations. The uncertainty related to the use of short records is important when the record length is comparable to the return period ( ∼ 50, ∼ 100, and ∼ 150 % for Mediterranean, semiarid, and arid climates, respectively). The agreement between IDF curves derived from different sensors on Mediterranean and, to a good extent, semiarid climates, demonstrates the potential of remote sensing datasets and instils confidence on their quantitative use for ungauged areas of the Earth.
The quantification of spatial rainfall is critical for distributed hydrological modeling. Rainfall spatial patterns generated by similar weather conditions can be extremely diverse. This variability can have a significant impact on hydrological processes. Stochastic simulation allows generating multiple realizations of spatial rainfall or filling missing data. The simulated data can then be used as input for numerical models to study the uncertainty on hydrological forecasts. In this paper, we use the direct sampling technique to generate stochastic simulations of high-resolution (1 km) daily rainfall fields, conditioned by elevation and weather state. The technique associates historical radar estimates to variables describing the daily weather conditions, such as the rainfall type and mean intensity, and selects radar images accordingly to form a con- ditional training image set of each day. Rainfall fields are then generated by resampling pixels from these images. The simulation at each location is conditioned by neighbor patterns of rainfall amount and eleva- tion. The technique is tested on the simulation of daily rainfall amount for the eastern Mediterranean. The results show that it can generate realistic rainfall fields for different weather types, preserving the temporal weather pattern, the spatial features, and the complex relation with elevation. The concept of conditional training image provides added value to multiple-point simulation techniques dealing with extremely non- stationary heterogeneities and extensive data sets.
Intensity–Duration–Frequency (IDF) curves are widely used to quantify the probability of occurrence of rainfall extremes. The usual rain gauge based approach provides accurate curves for a specific location, but uncertainties arise when ungauged regions are examined or catchment scale information is required. Remotely sensed rainfall records, e.g. from weather radars and satellites, are recently becoming available, providing high resolution information on rainfall extremes at regional or even global scales: their uncertainty and implications on water resources applications urge to be investigated. This study compares IDF curves from radar and satellite (CMORPH) estimates over the Eastern Mediterranean (covering Mediterranean, semiarid and arid climates) and quantifies the uncertainty related to their limited record on varying climates. We show that radar identifies thicker tail distributions than satellite, in particular for short durations, and that the shape parameters depends on the spatial and temporal aggregation scales. The spatial correlation between radar-IDFs and satellite-IDFs is as high as 0.7 for 2–5 years return period and decreases with longer return periods, especially for short durations. The uncertainty related to the use of short records is important when the record length is comparable to the return period (\~ 50 %, \~ 100 % and \~ 150 % for Mediterranean, semiarid and arid climates, respectively). The agreement between IDF curves derived from different sensors on Mediterranean and, to a good extent, semiarid climates, demonstrates the potential of remote sensing datasets and instils confidence on their quantitative use for ungauged areas of the Earth.
In this study we investigate the scaling of precipitation extremes with temperature in the Mediterranean region by assessing against observations the present day and future regional climate simulations performed in the frame of the HyMeX and MED-CORDEX programs. Over the 1979–2008 period, despite differences in quantitative precipitation simulation across the various models, the change in precipitation extremes with respect to temperature is robust and consistent. The spatial variability of the temperature–precipitation extremes relationship displays a hook shape across the Mediterranean, with negative slope at high temperatures and a slope following Clausius–Clapeyron (CC)-scaling at low temperatures. The temperature at which the slope of the temperature–precipitation extreme relation sharply changes (or temperature break), ranges from about 20 \$\backslash$textdegree\C in the western Mediterranean to \textless10 \$\backslash$textdegree\C in Greece. In addition, this slope is always negative in the arid regions of the Mediterranean. The scaling of the simulated precipitation extremes is insensitive to ocean–atmosphere coupling, while it depends very weakly on the resolution at high temperatures for short precipitation accumulation times. In future climate scenario simulations covering the 2070–2100 period, the temperature break shifts to higher temperatures by a value which is on average the mean regional temperature change due to global warming. The slope of the simulated future temperature–precipitation extremes relationship is close to CC-scaling at temperatures below the temperature break, while at high temperatures, the negative slope is close, but somewhat flatter or steeper, than in the current climate depending on the model. Overall, models predict more intense precipitation extremes in the future. Adjusting the temperature–precipitation extremes relationship in the present climate using the CC law and the temperature shift in the future allows the recovery of the temperature–precipitation extremes relationship in the future climate. This implies negligible regional changes of relative humidity in the future despite the large warming and drying over the Mediterranean. This suggests that the Mediterranean Sea is the primary source of moisture which counteracts the drying and warming impacts on relative humidity in parts of the Mediterranean region.
This work evaluates two numerical warning indicators of severe weather. These indicators, the MKI and RDI indices, were developed within the framework of the EU- funded FLASH project which studies flash flood events in the Mediterranean Basin. The MKI (Modified K-Index) is a modification of the K-Index, which expresses probability of lightning activity, and the RDI (Rain Dynamical Index) is the integrated upward moisture flux. The indices were tested on 59 episodes which occurred during nine rainstorms in Israel, Greece, Spain, Italy, and Cyprus. The data for calculation of the indices included rain cell identification derived from microwave radiometer imagery of polar orbiting NOAA satellites, rain RADAR data, and lightning activity from the international ZEUS detection system. Atmospheric data with 0.5? 9 0.5? spatial resolution and 6-h time res- olution were used for the calculation and the display of the two indices. The indices were tested by calculating the spatially correlating locations with high index values and actual locations of intense rain and intense lightning. The RDI detected both event types: rain and lightning, with a statistically significant success rate and a low rate of false results. The MKI was successful in indicating intense lightning activity, but the rate of correct indi- cations was not statistically significant and there was a high rate of false indications. The results suggest that the RDI computed with output of weather prediction models is a potentially good predictor of torrential rain and therefore can predict flash floods caused by such rain in the Mediterranean region.
The Dead Sea region has faced substantial environmental challenges in recent decades, including water resource scarcity, \~ 1 m annual decreases in the water level, sinkhole development, ascending-brine freshwater pollution, and seismic disturbance risks. Natural processes are significantly affected by human interference as well as by climate change and tectonic developments over the long term. To get a deep understanding of processes and their interactions, innovative scientific approaches that integrate disciplinary research and education are required. The research project DESERVE (Helmholtz Virtual Institute Dead Sea Research Venue) addresses these challenges in an interdisciplinary approach that includes geophysics, hydrology, and meteorology. The project is implemented by a consortium of scientific institutions in neighboring countries of the Dead Sea (Israel, Jordan, Palestine Territories) and participating German Helmholtz Centres (KIT, GFZ, UFZ). A new monitoring network of meteorological, hydrological, and seismic/geodynamic stations has been established, and extensive field research and numerical simulations have been undertaken. For the first time, innovative measurement and modeling techniques have been applied to the extreme conditions of the Dead Sea and its surroundings. The preliminary results show the potential of these methods. First time ever performed eddy covariance measurements give insight into the governing factors of Dead Sea evaporation. High-resolution bathymetric investigations reveal a strong correlation between submarine springs and neo-tectonic patterns. Based on detailed studies of stratigraphy and borehole information, the extension of the subsurface drainage basin of the Dead Sea is now reliably estimated. Originality has been achieved in monitoring flash floods in an arid basin at its outlet and simultaneously in tributaries, supplemented by spatio-temporal rainfall data. Low-altitude, high resolution photogrammetry, allied to satellite image analysis and to geophysical surveys (e.g. shear-wave reflections) has enabled a more detailed characterization of sinkhole morphology and temporal development and the possible subsurface controls thereon. All the above listed efforts and scientific results take place with the interdisciplinary education of young scientists. They are invited to attend joint thematic workshops and winter schools as well as to participate in field experiments
In this study we investigate the scaling of precipitation extremes with temperature in the Mediterranean region by assessing against observations the present day and future regional climate simulations performed in the frame of the HyMeX and MED-CORDEX programs. Over the 1979–2008 period, despite differences in quantitative precipitation simulation across the various models, the change in precipitation extremes with respect to temperature is robust and consistent. The spatial variability of the temperature–precipitation extremes relationship displays a hook shape across the Mediterranean, with negative slope at high temperatures and a slope following Clausius–Clapeyron (CC)-scaling at low temperatures. The temperature at which the slope of the temperature–precipitation extreme relation sharply changes (or temperature break), ranges from about 20 °C in the western Mediterranean to \textless10 °C in Greece. In addition, this slope is always negative in the arid regions of the Mediterranean. The scaling of the simulated precipitation extremes is insensitive to ocean–atmosphere coupling, while it depends very weakly on the resolution at high temperatures for short precipitation accumulation times. In future climate scenario simulations covering the 2070–2100 period, the temperature break shifts to higher temperatures by a value which is on average the mean regional temperature change due to global warming. The slope of the simulated future temperature–precipitation extremes relationship is close to CC-scaling at temperatures below the temperature break, while at high temperatures, the negative slope is close, but somewhat flatter or steeper, than in the current climate depending on the model. Overall, models predict more intense precipitation extremes in the future. Adjusting the temperature–precipitation extremes relationship in the present climate using the CC law and the temperature shift in the future allows the recovery of the temperature–precipitation extremes relationship in the future climate. This implies negligible regional changes of relative humidity in the future despite the large warming and drying over the Mediterranean. This suggests that the Mediterranean Sea is the primary source of moisture which counteracts the drying and warming impacts on relative humidity in parts of the Mediterranean region.
The typical short generation length of insects makes their population dynamics highly sensitive not only to mean annual temperatures but also to their intra-annual variations. To consider the combined effect of both thermal factors under global warming, we propose a modeling framework that links general circulation models (GCMs) with a stochastic weather generator and population dynamics models to predict species population responses to inter- and intra-annual temperature changes. This framework was utilized to explore future changes in populations of Bemisia tabaci, an invasive insect pest-species that affects multiple agricultural systems in the Mediterranean region. We considered three locations representing different pest status and climatic conditions: Montpellier (France), Seville (Spain), and Beit-Jamal (Israel). We produced ensembles of local daily temperature realizations representing current and future (mid-21st century) climatic conditions under two emission scenarios for the three locations. Our simulations predicted a significant increase in the average number of annual generations and in population size, and a significant lengthening of the growing season in all three locations. A negative effect was found only in Seville for the summer season, where future temperatures lead to a reduction in population size. High variability in population size was observed between years with similar annual mean temperatures, suggesting a strong effect of intra-annual temperature variation. Critical periods were from late spring to late summer in Montpellier and from late winter to early summer in Seville and Beit-Jamal. Although our analysis suggested that earlier seasonal activity does not necessarily lead to increased populations load unless an additional generation is produced, it is highly likely that the insect will become a significant pest of open-fields at Mediterranean latitudes above 40° during the next 50 years. Our simulations also implied that current predictions based on mean temperature anomalies are relatively conservative and it is better to apply stochastic tools to resolve complex responses to climate change while taking natural variability into account. In summary, we propose a modeling framework capable of determining distinct intra-annual temperature patterns leading to large or small population sizes, for pest risk assessment and management planning of both natural and agricultural ecosystems.
A modeling framework is formulated and applied to assess the sensitivity of the hydrological regime of two catchments in a convective rainfall environment with respect to projected climate change. The study uses likely rainfall scenarios with high spatiotemporal resolution that are dependent on projected changes in the driving regional meteorological synoptic systems. The framework was applied to a case study in two medium-sized Mediterranean catchments in Israel, affected by convective rainfall, by combining the HiReS-WG rainfall generator and the SAC-SMA hydrological model. The projected climate change impact on the hydrological regime was examined for the RCP4.5 and RCP8.5 emission scenarios, comparing the historical (beginning of the 21st century) and future (mid-21st-century) periods from three general circulation model simulations available from CMIP5. Focusing on changes in the occurrence frequency of regional synoptic systems and their impact on rainfall and streamflow patterns, we find that the mean annual rainfall over the catchments is projected to be reduced by 15% (outer range 2–23%) and 18% (7–25%) for the RCP4.5 sand RCP8.5 emission scenarios, respectively. The mean annual streamflow volumes are projected to be reduced by 45% (10–60%) and 47% (16–66%). The average events' streamflow volumes for a given event rainfall depth are projected to be lower by a factor of 1.4–2.1. Moreover, the streamflow season in these ephemeral streams is projected to be shorter by 22% and 26–28% for the RCP4.5 and RCP8.5, respectively. The amplification in reduction of streamflow volumes relative to rainfall amounts is related to the projected reduction in soil moisture, as a result of fewer rainfall events and longer dry spells between rainfall events during the wet season. The dominant factors for the projected reduction in rainfall amount were the reduction in occurrence of wet synoptic systems and the shortening of the wet synoptic systems durations. Changes in the occurrence frequency of the two dominant types of the regional wet synoptic systems (active Red Sea trough and Mediterranean low) were found to have a minor impact on the total rainfall.
Intensity–Duration–Frequency (IDF) curves are widely used in flood risk management because they provide an easy link between the characteristics of a rainfall event and the probability of its occurrence. Weather radars provide distributed rainfall estimates with high spatial and temporal resolutions and overcome the scarce representativeness of point-based rainfall for regions characterized by large gradients in rainfall climatology. This work explores the use of radar quantitative precipitation estimation (QPE) for the identification of IDF curves over a region with steep climatic transitions (Israel) using a unique radar data record (23yr) and combined physical and empirical adjustment of the radar data. IDF relationships were derived by fitting a generalized extreme value distribution to the annual maximum series for durations of 20min, 1h and 4h. Arid, semi-arid and Mediterranean climates were explored using 14 study cases. IDF curves derived from the study rain gauges were compared to those derived from radar and from nearby rain gauges characterized by similar climatology, taking into account the uncertainty linked with the fitting technique. Radar annual maxima and IDF curves were generally overestimated but in 70% of the cases (60% for a 100yr return period), they lay within the rain gauge IDF confidence intervals. Overestimation tended to increase with return period, and this effect was enhanced in arid climates. This was mainly associated with radar estimation uncertainty, even if other effects, such as rain gauge temporal resolution, cannot be neglected. Climatological classification remained meaningful for the analysis of rainfall extremes and radar was able to discern climatology from rainfall frequency analysis.
The effect of climate change on the Eastern Mediterranean (EM) region, a region that reflects a transition between Mediterranean and semi-arid climates, was examined. This transition region is affected by global changes such as the expansion of the Hadley cell, which leads to a poleward shift of the subtropical dry zone. The Hadley cell expansion forces the migration of jet streams and storm tracks poleward from their standard course, potentially increasing regional desertification. This article focuses on the northern coastline of Israel along the EM region where most wet synoptic systems (i.e. systems that may lead to precipitation) are generated. The current climate was compared to the predicted mid-21st century climate based on Intergovernmental Panel on Climate Change (IPCC) Representative Concentration Pathway (RCP) RCP4.5 and RCP8.5 scenarios using four Coupled Model Intercomparison Project Phase 5 (CMIP5) models. A warming of 1.1–2.6 °C was predicted for this region. The models predicted that rain in the region will become less frequent, with a reduction of 1.2–3.4% in 6-h intervals classified as wet synoptic systems and a 10–22% reduction in wet events. They further predicted that the maximum wet event duration in the mid-21st century would become shorter relative to the current climate, implying that extremely long wet systems will become less frequent. Three of the models predicted shrinking of the wet season length by up to 15%. All models predicted an increasing occurrence frequency of Active Red Sea Troughs (ARSTs) for the RCP8.5 scenario by up to 11% by the mid-21st century. For the RCP4.5 scenario, a similar increase of up to 6% was predicted by two of the models.
Prolonged dry spells (PDSs) during the rainy season have severe environmental implications, including water shortage, damage to agriculture and increased potential for forest fires. This holds in particular for vulnerable regions, such as the Levant, already subjected to decrease in rainfall and lengthening of dry spells, in agreement with predictions of climatic models for the coming decades. This is the first comprehensive study which identifies atmospheric patterns responsible for PDS occurrence on thousands of kilometres scale. A total of 178 PDSs, of \textgreater7 days, were found within the 62 seasons studied. A subjective inspection of upper-level geopotential height (GPH), sea-level pressure (SLP) and lower-level temperature anomalies point at three types, each associated with a definite climatic regime. The ‘subtropical' type is associated with an expansion of the subtropical high over the majority of the Mediterranean, accompanied by northward migration of the Mediterranean cyclone track. The ‘baroclinic', the most frequent type, is induced by a pronounced stagnant ridge over the eastern Mediterranean, being a part of Rossby wave, accompanied by a pronounced trough/cut-off low over the western Mediterranean. The ‘polar' type results from intrusion of lower-level continental polar air associated with upper-level trough east of the Levant and blocking high over central Europe. Quantitative indices were derived for objective classification of the types, based on the climatic regimes defined subjectively, and the centers of action representing each. Composite maps for each type indicate substantial differences in the synoptic configuration and the factors explaining absence of rain. For the subtropical type, the dynamic factor of subsidence is dominant. For the polar, the thermodynamic factor of continental dry advection is dominant and for the baroclinic, both dynamic and thermodynamic factors are important. Classification of PDSs according to synoptic scenarios enables analysis of future changes in the occurrence and duration pattern of PDSs, using output of climate models.
This work analyses the prominent characteristics of extreme storms and flash-flood regimes in two main areas of the Mediterranean region: the North-Western (comprising Spain, France and Italy) and South-Eastern region (Israel). The two areas are chosen to represent the two end members of variation in flash-flood regimes in the Mediterranean basin. Data from 99 events collected in the two areas (69 from the North-Western region and 30 from the South-Eastern region), for which occurrence date, catchment area and flood peak are available, were used to provide a detailed description the flash-flood seasonality patterns, the synoptic and mesoscale atmospheric controls, and flood envelope relationship. Results show that the flood envelope curve for the South-Eastern region exhibits a more pronounced decreasing with catchment size with respect to the curve of the North-Western region. The differences between the two relationships reflect variations in the fractional storm coverage of the basin and hydrological characteristics between the two regions. Seasonality analysis shows that the events in the North-Western region tend to occur between August and November, whereas those in the South-Eastern area tend to occur in the period between October and May, reflecting the relevant patterns in the synoptic conditions controlling the generation of intense precipitation events.
Abstractt Spatio-temporal storm properties have a large impact on catchment hydrological response. The sensitivity of simulated flash floods to convective rain-cell characteristics is examined for an extreme storm event over a 94 km2 semi-arid catchment in southern Israel. High space–time resolution weather radar data were used to derive and model convective rain cells that then served as input into a hydrological model. Based on alterations of location, direction and speed of a major rain cell, identified as the flooding cell for this case, the impacts on catchment rainfall and generated flood were examined. Global sensitivity analysis was applied to identify the most important factors affecting the flash flood peak discharge at the catchment outlet. We found that the flood peak discharge could be increased three-fold by relatively small changes in rain-cell characteristics. We assessed that the maximum flash flood magnitude that this single rain cell can produce is 175 m3/s, and, taking into account the...