Next Article in Journal
Nitrogen Oxides (NOx) in the Arctic Troposphere at Ny-Ålesund (Svalbard Islands): Effects of Anthropogenic Pollution Sources
Next Article in Special Issue
A Modeling Study on the Downslope Wind of “Katevatos” in Greece and Implications for the Battle of Arachova in 1826
Previous Article in Journal
Toward Development of a Framework for Prediction System of Local-Scale Atmospheric Dispersion Based on a Coupling of LES-Database and On-Site Meteorological Observation
Previous Article in Special Issue
The Impact of the Variation in Weather and Season on WRF Dynamical Downscaling in the Pearl River Delta Region
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 12
Issue 7
10.3390/atmos12070902
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Evaluating the Forecast Skill of a Hydrometeorological Modelling System in Greece

by
George Varlas
*,
Anastasios Papadopoulos
,
George Papaioannou
and
Elias Dimitriou
Institute of Marine Biological Resources and Inland Waters, Hellenic Centre for Marine Research, Anavyssos, 19013 Attica, Greece
*
Author to whom correspondence should be addressed.
Atmosphere 2021, 12(7), 902; https://doi.org/10.3390/atmos12070902
Submission received: 15 June 2021 / Revised: 5 July 2021 / Accepted: 9 July 2021 / Published: 13 July 2021
(This article belongs to the Special Issue Advances in Mesoscale Numerical Weather Prediction and its Applications)
Browse Figures
Versions Notes

Abstract

:
A hydrometeorological forecasting system has been operating at the Institute of Marine Biological Resources and Inland Waters (IMBRIW) of the Hellenic Centre for Marine Research (HCMR) since September 2015. The system consists of the Advanced Weather Research and Forecasting (WRF-ARW) model, the WRF-Hydro hydrological model, and the HEC-RAS hydraulic–hydrodynamic model. The system provides daily 120 h weather forecasts focusing on Greece (4 km horizontal resolution) and hydrological forecasts for the Spercheios and Evrotas rivers in Greece (100 m horizontal resolution), also providing flash flood inundation forecasts when needed (5 m horizontal resolution). The main aim of this study is to evaluate precipitation forecasts produced in a 4-year period (September 2015–August 2019) using measurements from meteorological stations across Greece. Water level forecasts for the Evrotas and Spercheios rivers were also evaluated using measurements from hydrological stations operated by the IMBRIW. Moreover, the forecast skill of the chained meteorological–hydrological–hydraulic operation of the system was investigated during a catastrophic flash flood in the Evrotas river. The results indicated that the system provided skillful precipitation and water level forecasts. The best evaluation results were yielded during rainy periods. They also demonstrated that timely flash flood forecasting products could benefit flood warning and emergency responses due to their efficiency and increased lead time.

1. Introduction

The interconnected physical processes between the atmosphere and hydrosphere affect the water cycle of the planet and sometimes trigger severe hydrometeorological phenomena like floods [1]. Flooding is one of the most frequent natural hazards often posing a threat to human lives and structural utilities. As floods present an overall increasing trend worldwide during the three last decades [2], there is the necessity to increase the preparedness and the protection of socio-economic activities and human lives. Given the fact that no preventative measures or defense structures can be completely effective, flood-risk management systems should provide as much sufficient lead time as feasible. Therefore, the implementation and operation of flood forecasting and warning systems are important parts of flood management services [3,4,5].
In this context, several operational flood forecasting systems have been implemented, especially during the last two decades. These can be divided into two main categories, the global-scale and the continental-scale flood forecasting systems [6]. As regards the global-scale flood forecasting systems, the Global Flood Awareness System (GloFAS [7]) is one of the most popular. Since July 2011, GloFAS has been producing global ensemble streamflow forecasts and flood early warnings. In addition, GloFAS was recently upgraded to produce seasonal hydrological forecasts [8]. As regards the continental-scale flood forecasting systems, the National Water Model (NWM) and the European Flood Alert System (EFAS) are two of the most popular in the United States and Europe, respectively. The NWM system is a hydrological modelling framework providing forecasts from 18 h to 30 days in advance. The NWM has been operated by the National Oceanic and Atmospheric Administration (NOAA) over the entire continental United States since August 2016 [9]. The core of the NWM system is the Weather Research and Forecasting Hydrological model (WRF-Hydro [10]) set up on a 250 m × 250 m grid. WRF-Hydro has been developed by the National Center for Atmospheric Research (NCAR), also including community-based development processes. Regarding input data, it assimilates streamflow information from hydrological stations while ingesting meteorological forecasting data from a variety of sources such as the Global Forecast System (GFS) of the National Centers for Environmental Prediction (NCEP). A detailed description of the NWM system is online available at: https://water.noaa.gov/about/nwm (accessed on 12 July 2021). The EFAS system has been performing probabilistic hydrological forecasts at the Joint Research Centre (JRC) of the European Commission since 2005 [11]. It aims at providing river flood forecasts to hydrometeorological services in European countries so as to increase their preparedness. EFAS is set up on a 5 km × 5 km grid and provides forecasts from 4 h to 8 weeks in advance. EFAS uses historical and real-time hydrometeorological measurements to approximate flood thresholds and to define initial hydrological conditions, respectively. As regards meteorological information, EFAS uses weather forecasts from the German Weather Service (DWD) and the European Centre for Medium-Range Weather Forecasts (ECMWF) as well as 51 probabilistic forecasts from the Ensemble Prediction System (EPS) of the ECMWF [12].
In the Mediterranean area, catastrophic flash floods are usually characterized by their small spatiotemporal scales [13,14,15,16]. Greece also suffers from local flash floods causing severe damages, economic losses, and even fatalities [17]. Noteworthy flash flood events in Greece occurred at Mandra town on 15 November 2017 [17,18,19,20,21,22,23], at Volos city on 9 October 2006 [24], and at Rafina catchment on 11 December 2009 and 22 February 2013 [25]. The hydrometeorological forecasting and warning services in small-scale catchments demand a high resolution to capture the specific spatiotemporal characteristics of the flash flood events [26]. However, operational small-scale flash flood forecasting systems still require a considerable effort in research and development. Recently, flash flood modelling in small-scale basins of Greece has usually been based on high-resolution hydrological models forced by precipitation estimates using nowcasting approaches (e.g., [22]) regarding short-term (e.g., 1–3 h in advance) forecasts or by numerical weather prediction results (e.g., [17,18,24]) regarding forecasts exceeding even 1–2 days in advance.
In this context, the Institute of Marine Biological Resources and Inland Waters (IMBRIW) of the Hellenic Centre for Marine Research (HCMR) has been operating and continuously upgrading a high-resolution hydrometeorological forecasting system since September 2015. The forecasts are online available from the website: https://meteo.hcmr.gr (accessed on 12 July 2021). This study aims at assessing the forecast skill of the IMBRIW’s hydrometeorological system over Greece for the 4-year period from September 2015 to August 2019. Precipitation forecasts were evaluated using measurements from land surface meteorological stations across Greece included in the network of the World Meteorological Organization (WMO). Moreover, water level forecasts in the Evrotas and Spercheios rivers were evaluated using water level measurements from hydrological stations operated by the IMBRIW. The hydrological forecast skill was also investigated during a flash flood in the Evrotas river that occurred on 7 September 2016 and caused one fatality, severe damages, and the overflow of a bridge at Skala town.
The rest of the manuscript is organized as follows: In Section 2, an overview of the IMBRIW’s hydrometeorological modelling system is presented describing also the setup of each model. Section 3 presents the forecast skill evaluation methodology. Section 4 presents the 4-year evaluation results regarding precipitation and river water level as well as the investigation of forecast skill in the flash flood event at the Evrotas river. The results are discussed and interpreted in Section 4. Finally, Section 5 contains the conclusions of this work.

2. Overview of the Hydrometeorological System

The IMBRIW’s hydrometeorological system consists of atmospheric, hydrological, and hydraulic–hydrodynamic models. It is noteworthy that the integration of interdisciplinary modelling approaches better facilitates the representation and prediction of water and energy cycles, increasing the physical content of forecasting products. In particular, the system consists of the Advanced Weather Research and Forecasting (WRF-ARW) model [27], the WRF-Hydro hydrological model [10], and the HEC-RAS hydraulic–hydrodynamic model [28]. The WRF-ARW model is set up to daily produce 120-h weather forecasts for three nested domains at different horizontal resolutions (36, 12, and 4 km) across the Mediterranean basin, focusing on Greece (4 km). Respectively, WRF-Hydro is set up to daily produce 120-h streamflow and water level forecasts for two river basins in Greece, i.e., the Spercheios and Evrotas river basins. It is important to note that, the capability to provide inundation forecasting using the HEC-RAS model forced by WRF-Hydro results is a recent enhancement of the system based on knowledge obtained during previous flash flood studies in Volos city [24] and Mandra town [18]. Therefore, this study only presents the application of HEC-RAS in the flash flood event that occurred on 7 September 2016 at the Evrotas river overflowing a bridge at Skala town. Figure 1 presents a flowchart which summarizes the main data used and model setup as described in the following sub-sections.

2.1. Atmospheric Model Setup

As regards the atmospheric component of the system, the WRF-ARW model has been installed and appropriately configured on the parallel computing infrastructure of the Laboratory of Hydrometeorology (LHM) of the IMBRIW-HCMR. In particular, version 3.7 of the WRF-ARW model [27] was used during the evaluation period from September 2015 to August 2019. As shown in Figure 2a, three nested domains were considered: (D1) the coarse domain with 36 km × 36 km horizontal resolution (161 × 90 Arakawa-C grid points), (D2) the fine domain with 12 km × 12 km horizontal resolution (211 × 160 Arakawa-C grid points), and (D3) the finest domain with 4 km × 4 km horizontal resolution (199 × 199 Arakawa-C grid points). The vertical structure of all domains consisted of 38 levels up to a 50 hPa pressure top (about 20 km). The WRF-ARW model runs every day based on initial and boundary conditions obtained from the operational analyses at 12:00 UTC of the Global Forecasting System (GFS) of the National Centers for Environmental Prediction (NCEP) in the horizontal resolution of 0.5° × 0.5°. For the lower boundary conditions over the sea, sea surface temperature (SST) was refreshed daily preserving it constant throughout the 5-day simulations. During the evaluation period, initial SST fields were based on the high-resolution (0.083° × 0.083°) real time global (RTG) SST analyses produced by the NCEP. Furthermore, land use and topographic data were obtained from the United States Geological Survey (USGS) database. As far as the parameterization schemes are concerned, many tests were performed before September 2015 to find which of them resulted in realistic simulations of weather conditions over the area covered. Hence, the revised Monin–Obukhov scheme [29] was used during the evaluation period to represent the processes in the atmospheric surface layer. For the simulation of planetary boundary layer processes, the Yonsei University scheme [30] was used. The land surface and soil processes were simulated using the Unified Noah scheme [31]. In order to resolve the longwave and shortwave radiation processes the RRTMG scheme [32] was used, while the Thompson scheme [33] was used for cloud microphysics. Moreover, the Grell–Freitas ensemble scheme [34] was used to parameterize the convective processes in the D1 and D2 domains while explicit resolve was applied in the D3 domain.

2.2. Hydrological Model Setup

In order to estimate streamflow and water level in the Spercheios and Evrotas rivers during the evaluation period, LHM applied the version 3.0 of the WRF-Hydro hydrological model [10]. The WRF-Hydro model was installed on the parallel computing infrastructure of the LHM and configured adopting an offline coupling approach using forcing fields provided by the WRF-ARW model. Forcing meteorological data used in the WRF-Hydro simulations were provided by the D3 (4 km × 4 km) domain of the WRF-ARW model. The forcing fields used are the following: liquid water precipitation rate, air temperature at 2 m, specific humidity at 2 m, incoming shortwave and longwave radiation, u- and v-components of wind at 10 m, and surface pressure [10]. As regards the areas covered by the hydrological simulations, the WRF-Hydro model was configured on two nested domains of WRF-ARW including Evrotas (73 × 73 Arakawa-C grid points) and Spercheios (73 × 37 Arakawa-C grid points) river basins, respectively. The nested domains used in the hydrological simulations were designed with horizontal resolution of 1 km × 1 km (Figure 2b,c). WRF-Hydro used the Noah land surface model (LSM) [35] in the 1 km × 1 km horizontal resolution to represent land processes, but routing processes were represented in a 10-times finer horizontal resolution (100 m × 100 m), applying aggregation/disaggregation procedures [10]. The WRF-Hydro model estimates infiltration and exfiltration using a diffusive wave overland routing scheme [36,37] while it calculates channel routing using the Muskingum-Cunge approach [38].
Additionally, the Shuttle Radar Topographic Mission (SRTM) Digital Elevation Model (DEM) [39] data in the resolution of 90 m × 90 m distributed by the National Aeronautics and Space Administration (NASA) and the United States Geological Survey (USGS) was used to construct high-resolution topographic datasets used in WRF-Hydro simulations. More specifically, the void-filled version [40] of this data obtained from the Hydrological Data and Maps Based on Shuttle Elevation Derivatives at Multiple Scales (HydroSHEDS; https://hydrosheds.cr.usgs.gov/index.php (accessed on 12 July 2021) was used, regridding it to the 100 m × 100 m Evrotas and Spercheios domains (Figure 2b,c). These topographic datasets were also used to estimate stream order classification [41] for the two river basins, which was necessary for the hydrological simulations as it implies flow accumulation information. For the channel routing setup in the two rivers, the Manning roughness coefficient, the channel bottom width, and the slide slope were set for each stream order value as demonstrated in Table 1. These channel parameter values were selected performing many simulations before 1 September 2015 in adverse and mild weather conditions. Moreover, two important parameters adjusting the overland flow before water reaching in channels were set up. In more detail, the overland flow roughness scaling factor and the initial retention height scaling factor were set to 0.3 after extensive tests in the two rivers. Given the lack of continuous streamflow measurements in the two rivers, we exploited useful information derived from previous studies using the WRF-Hydro model in Greek catchments [17,18,22,24] as well as in catchments located at adjacent countries, such as Cyprus [42], Italy [43,44], and Turkey [45,46], to set the abovementioned calibration parameters so as to give reasonable results.

2.3. Hydraulic–Hydrodynamic Model Setup

The Hydrologic Engineering Center’s (CEIWR-HEC) River Analysis System (HEC-RAS) was used for river flood modelling and mapping. The free software HEC-RAS 6.0 version is capable of 2D unsteady flow simulation [47]. HEC-RAS 2D modelling and mapping had already been implemented successfully in several recent studies [18,24,48,49,50,51,52,53,54]. Moreover, some studies tested the recent capabilities of the updated HEC-RAS [55] and the efficiency of the two-dimensional (2D) HEC-RAS model [56]. Therefore, HEC-RAS 2D was selected as the most appropriate hydraulic–hydrodynamic model to be incorporated into the IMBRIW’s hydrometeorological system and, in this study, its performance is assessed for the catastrophic flash flood event of 7 September 2016 at the Evrotas river.
However, no conventional flood-related data (such as streamflow or water depth data from gauging stations close to the floodplain, and high-water marks of the most affected areas) exists for the study event. Thus, the hydraulic modelling setup and calibration were based on limited available non-conventional flood data (photographs and videos) collected from various sources (e.g., local mass-media reports). Therefore, the hydraulic modelling application was performed at the bridge’s location at Skala town (Figure 2b and Figure 3).
In 2D fluvial flood modelling, the Digital Elevation Model (DEM) accuracy is very important [57,58,59,60,61]; therefore, special care has been undertaken for the development of the area’s Digital Surface Model (DSM) and Digital Terrain Model (DTM). In particular, this procedure was based on high-resolution river geometry data created by processing Unmanned Aerial Vehicle (UAV) imagery (Figure 3a,b). The spatial resolution of the area’s DSM/DTM used was 11.521 cm.
The upstream boundary condition was defined using flood hydrograph (i.e., streamflow) forecast provided by the WRF-Hydro hydrological model, while the downstream boundary condition was determined as normal water depth or energy slope [28,47,62,63]. The bridge representation was based on data retrieved from the topographical survey contacted for the flood management risk plans of east Peloponnese (Figure 3c,e) [64]. The technical details of the bridge piers and abutments are presented in Figure 3d.
Based on the collected non-conventional flood data, i.e., photographs of Skala’s bridge taken during the flash flood event (Figure 3f) [65,66], we estimated the flood water depth to be approximately 6 m at the bridge location. According to this estimation, and following a typical trial and error optimization technique, the Manning roughness coefficient was set to 0.09.

3. Evaluation Methodology

The precipitation forecasts were evaluated for the 4-year period from 1 September 2015 to 31 August 2019 using precipitation data measured from 24 land surface meteorological stations over Greece. These stations are included in the network of the WMO and the measurements were available from the ECMWF. Except for precipitation, these stations measure additional conventional meteorological parameters such as temperature and dew point at 2 m, atmospheric pressure, wind speed, and direction at 10 m, etc. The locations of the 24 meteorological stations are illustrated in Figure 4. Additionally, Table 2 presents information for the 24 stations used in the evaluation of precipitation forecasts. Table 2 includes information about WMO identification code (WMO ID), the latitude, the longitude, and the altitude of each meteorological station over Greece.
The evaluation methodology was based on a point-to-point comparison between simulated and measured values [67]. In more detail, the simulated 6-h accumulated precipitation values were spatially co-located with the measured ones, using the four nearest neighboring points of the D3 model grid. Similar methodology was used by [68]. It is important to note here that the measurements of precipitation are available at 00:00 and 12:00 UTC (6-h accumulated values) and at 06:00 and 18:00 UTC (12-h accumulated values). In this study, we used 6-h accumulated precipitation values that resulted from a processing of these 6-h and 12-h values. To combine model results at the four points giving the nearest points higher impact, a form of inverse distance weighting interpolation was used. Considering that p1, p2, p3, and p4 are the simulated 6-h accumulated precipitation values at the four nearest neighboring points around the station location, as well as r1, r2, r3, and r4 are the respective distances between the points and the station location, the simulated 6-h accumulated precipitation value (p) used in the evaluation is estimated as follows
p = ∑ i = 1 4 p i r i ∑ i = 1 4 1 r i ,         i = 1 , 2 , 3 , 4
The evaluation scores used for the 6-h accumulated precipitation values were based on the contingency table approach [69]. As shown in Table 3, the contingency table is a two-dimensional matrix in which each element counts the number of measurements and model forecasts exceeding or failing to reach a certain threshold. The table elements are defined as follows:
  • A: model forecast and measurement exceeded the threshold.
  • B: model forecast exceeded the threshold but measurement did not.
  • C: model forecast did not reach the threshold but measurement exceeded it.
  • D: model forecast and measurement did not reach the threshold.
Considering the table elements, the precipitation forecast skill was evaluated, estimating the bias score (BIAS) and the equitable threat score (ETS), using pairs of simulated and measured 6-h accumulated precipitation values while considering 11 precipitation thresholds (1, 3, 6, 9, 12, 15, 18, 21, 24, 27, and 30 mm).
Thus, the BIAS is defined as
BIAS = A + B A + C  
while the ETS is defined as
ETS = A − E A + B + C − E  
where E, a term to remove the number of random correct forecasts, is defined by
E = A + B × A + C N
with N representing the total number of measurements (i.e., N = A + B + C + D) [67]. At given precipitation thresholds, the BIAS may yield a systematic overestimation (when BIAS > 1) or underestimation (when BIAS < 1), and the ETS may present poor forecasts (when ETS ≈ 0) or perfect forecasts (when ETS = 1). More details about the BIAS and ETS scores are available in [67,70].
The evaluation procedure for water level was also based on a point-to-point comparison between daily average simulated values and daily average measured values from the Sentenikos and Anthili hydrological stations of IMBRIW in the Evrotas and Spercheios rivers, respectively (Figure 2b,c). The measurements used in this study covered a period from 8 April 2016 to 31 August 2019 regarding the Sentenikos station (Figure 2b) and a period from 4 September 2015 to 31 August 2019 regarding the Anthili station (Figure 2c). It is important to mention here that IMBRIW operates a network of hydrological stations covering various Greek catchments measuring water level among other parameters [71], while it also archives measurements in a database. The simulated water level values were averaged daily and spatially co-located with the daily average measured values. For the spatial co-location, the nearest to the stations neighboring river points of the 100 m × 100 m hydrological model grids were used. Consequently, the water level forecast skill was evaluated estimating the mean bias error (MBE [69]), the root mean square error (RMSE [69]), and the Nash–Sutcliffe efficiency (NSE [72]). The formulas of these evaluation scores are shown below
MBE = ∑ i = 1 N S i − M i N
RMSE = ∑ i = 1 N S i − M i 2 N
NSE = 1 − ∑ i = 1 N S i − M i 2 ∑ i = 1 N M i − M ¯ 2
where Si and Mi are the daily average simulated and measured values, respectively, M ¯ is the mean of measured values, and N is the total number of measurements.

4. Results and Discussion

This section presents the results of forecast skill evaluation regarding precipitation over Greece and water level in the Evrotas and Spercheios rivers. Specific forecast skill characteristics of the atmospheric and hydrological models are revealed. Afterwards, this section presents the tri-model (i.e., meteorological, hydrological, and hydraulic–hydrodynamic) forecast of the Evrotas river flash flood event that occurred during a severe storm in the morning of 7 September 2016. The results of the sequential simulations are analyzed to investigate the operational forecast skill of the integrated modelling system even in such adverse weather conditions.

4.1. Evaluation of Precipitation and River Water Level Forecast Skill for the 4-Year Period

Figure 5 demonstrates the forecast skill of the meteorological model regarding 6-h accumulated precipitation. The forecast skill is expressed by the dependence of BIAS and ETS evaluation scores to the forecast horizon up to 120 forecast hours (FH) ahead. It is important to note that only 6-h accumulated precipitation values exceeding the threshold of 1 mm were considered to show the relation between the forecast skill and the forecast horizon in weak, moderate, and heavy precipitation conditions. The BIAS score ranges from 0.73 at the 6th FH to 1.24 at the 108th FH and it is characterized by variations around one as the forecast horizon increases, implying alternations of overestimation (higher than one) and underestimation (lower than one). BIAS alternations seem to follow an afternoon–evening–morning–noon pattern during the 5-day forecast horizon implying a dependence of precipitation forecast on semi-diurnal cycles. The worst BIAS (i.e., 0.73) is presented in the first 6 FH and it is related with the reduced predictability of the model during its spinup period. Overall, the BIAS score does not present systematic trends with the forecast horizon. On the other hand, the ETS presents an increase in the first 18 FH, reaching a value of 0.38, and then it gradually reduces to 0.15. Comparable ETS values and similar linear ETS reduction with the forecast horizon have also been documented by previous studies (e.g., [67,73,74]).
Figure 6 shows the dependence of BIAS and ETS on the 11 thresholds of 6-h accumulated precipitation used in the evaluation procedure. The diagrams refer to all the 24 meteorological stations and the forecasts for the first day (i.e., from the 12th to the 36th FH) were considered, excluding the first 12 FHs to avoid spinup issues. Figure 6a shows very good BIAS equal to 0.98 regarding 6-h accumulated precipitation that exceeds 1 mm, thus including almost all of the precipitation intensities (weak, moderate, and heavy). However, considering higher precipitation thresholds, BIAS ranges from 0.83 to 0.65 implying underestimation is more pronounced for heavy precipitation. Respectively, ETS ranges from 0.34 to 0.03 with the ETS values worsening for the most intense precipitation thresholds (>27 mm/6-h) (Figure 6b).
A seasonal analysis of precipitation forecast skill is presented in Figure 7. In more detail, Figure 7a–d shows BIAS for winter, spring, summer, and autumn, respectively, while Figure 7e–h demonstrates ETS for the same seasons. Similarly to the overall BIAS and ETS of Figure 6, 6-h accumulated precipitation values considering the 11 thresholds were used in the evaluation procedure. Precipitation underestimation and reduction of ETS with forecast horizon are evidenced in all the seasons. Nevertheless, there are seasonal variabilities of both BIAS and ETS with the worst forecast skill in summer and the best in autumn and winter. In spring, BIAS is better compared to the other seasons but ETS is worse than the others in winter and autumn. It is noteworthy that an improvement in both scores is noted in summer regarding high thresholds. This may be explained by the explicit resolve of convection used by the model, thus resulting in sufficient representation of physical processes during the formation of summer thunderstorms, which usually produce heavy precipitation.
Figure 8a,b illustrates all the 24 stations included in the evaluation of the model colored to demonstrate precipitation BIAS and ETS, respectively. Similarly to Figure 6, the 6-h accumulated precipitation (mm) amounts over 1 mm were considered in the estimation of the contingency table for each station. There are spatial variabilities of BIAS and ETS scores that are attributed to the local climate of each area studied. Even though it is difficult to draw robust conclusions for the forecast skill due to the blended BIAS and ETS values presented across Greece, it is interesting to note the systematic underestimation (BIAS lower than one) illustrated at stations in western Greece. This may be attributed to the weakness of the model to represent large amounts of precipitation during the passage of barometric lows and fronts that originated from western sea areas. This weakness is partly explained by the lack of sea variabilities representation in the model forecasts. This could be feasible by implementing an air–sea coupled forecasting system “online” representing air–sea processes and feedbacks in the surface layer calculations of the meteorological model [75,76,77,78,79,80]. Such modelling systems not only improve the representation of surface fluxes but also simulate more realistically water cycle processes resulting in improvements of precipitation forecasts [17,18].
Figure 9a,b demonstrates timeseries of the simulated and measured water level at the Sentenikos station in the Evrotas river and the Anthili station in the Spercheios river, respectively. The timeseries cover periods from 4 September 2015 to 31 August 2019 and from 8 April 2016 to 31 August 2019 at the Anthili and Sentenikos stations, respectively. Regarding the simulated values, forecasts of daily average water level referring to the first day (i.e., from the 12th to the 36th FH) were used in the evaluation and are presented in Figure 9a,b. As shown in Figure 9a,b, the hydrological model sufficiently captures the water level variations at both river locations following the rainy-period peaks and yielding reasonable statistical scores. Regarding the Sentenikos station, MBE, RMSE, and NSE are equal to −3.52, 16.59, and 0.36, respectively. The scores are better at the Anthili station, reaching 0.15, 13.09, and 0.41, respectively. This may be attributed to the downstream location of the Anthili station compared to the upstream location of the Sentenikos station, facilitating a more complete hydrological representation of the river basin.

4.2. Evaluation of Flash Flood Forecast Skill: The Case of 7 September 2016, Evrotas River

In an effort to investigate the flood forecast skill of the integrated system in its operational application, a flash flood in Evrotas river basin was selected to be studied. The flash flood occurred on 7 September 2016 and resulted in one fatality, severe economic losses, and extended damages (transportation networks, buildings, and agricultural areas). The flash flood was triggered by a severe storm over the Evrotas river basin. As shown in Figure 10a, a cut-off trough over the central Mediterranean Sea supported southern atmospheric circulation over the study area on 7 September at 06:00 UTC. The southern circulation transferred warm and moist air from sea areas to the mountainous area around Evrotas. At the same time, the colder air in the center of the cut-off trough supported the formation of a cold front over western Greece (Figure 10b). This cold front swept through the Evrotas basin causing torrential rainfall, especially over the mountainous areas of the basin, which subsequently triggered the flash flood.
The results demonstrate that the atmospheric model was capable of reasonably simulating the storm causing the flash flood. As demonstrated in Figure 11, the meteorological model produced heavy 24-h precipitation (mm) from 6 September at 21:00 to 7 September at 21:00 UTC, a period that corresponds to daily precipitation on 7 September (i.e., from 7 September at 00:00 to 8 September at 00:00) if considering local time (UTC + 3 h) to be consistent with measurements. The model seems to underestimate the daily precipitation compared to measurements from meteorological stations across the Evrotas basin or at adjacent areas (Figure 11). In more detail, despite the fact that the model overestimated the 24-h precipitation at the Sparta (measurement: 67 mm and forecast: 85.6 mm) and the Molaoi (measurement: 38.2 mm and forecast: 61.3 mm) stations, it yielded noticeable underestimation at the Geraki (measurement: 200.2 mm and forecast: 49.3 mm) and the Kardamyli (measurement: 130 mm and forecast: 55.5 mm) stations (Figure 11). Moreover, the model resulted in a maximum 24-h accumulated precipitation of 160 mm at the west of the river basin, however, the maximum measured amount was 200.2 mm at the Geraki station located at the east of the basin.
Despite the quantitative and spatial inaccuracies of the forecast, the simulated precipitation over mountainous areas, especially for a 4-h period from 01:00 to 05:00 UTC, was enough to trigger the flash flood (Figure 12a–d). Subsequently, the WRF-Hydro hydrological model captured the river flooding characteristics during the period from 05:00 to 10:00 UTC (Figure 12d–i). The overall maximum simulated streamflow occurred at 07:00 UTC reaching 767 m3 s−1 (Figure 12f). The streamflow peak upstream from Skala’s bridge (location: 36.8602° N, 22.6677° E) occurred at 08:00 and 09:00 UTC reaching ~705 m3 s−1 (Table 4). Streamflow values resulting from the hydrological simulation at this location were used as input hydrograph in the hydraulic–hydrodynamic simulation in order to produce the inundation forecast at Skala’s bridge (Table 4).
Figure 13 presents the maximum water depth and the flood extent simulated by the 2D flood inundation model using the flood hydrograph generated from the WRF-Hydro hydrological model (Table 4). In addition, Figure 13b,c shows the maximum water depth profile for the Profile line (indicated by the green color in Figure 13a) that is upstream of the bridge and the timeseries of water depth at the deepest part of the profile line, respectively. According to the results presented in Figure 13, the simulated water depth converges to a maximum value of 6 m and there is overbank flow from both sides of the bridge. Indeed, qualitative analysis of the collected flood-related data (photographs and videos) at the Skala’s bridge location, in combination with the topographical survey sketch, indicated that the maximum water depth at the specific location was around or even higher than 6 m, creating flooding on both sides of the bridge.

5. Conclusions

This study assesses a hydrometeorological system operating since 2015 at the Institute of Marine Biological Resources and Inland Waters (IMBRIW) of the Hellenic Centre for Marine Research (HCMR). Precipitation forecasts produced for a 4-year period (September 2015–August 2019) were evaluated using measurements from 24 meteorological stations across Greece. Water level forecasts were also evaluated using measurements from two Greek hydrological stations of IMBRIW on the Evrotas and Spercheios rivers. Furthermore, the forecast skill of the system set up in a coupled meteorological–hydrological–hydraulic operation was investigated during a flash flood in Evrotas river (7 September 2016), causing one fatality, extensive damages, and overflow of a bridge at Skala town.
The results of the statistical evaluation indicated that the use of the hydrometeorological modelling system provides skillful precipitation and water level forecasts. The forecast skill evaluation for the 4-year period indicated that the modelling system is reliable during all of the seasons, showing better scores during the rainy periods of the year (i.e., autumn and winter). However, this finding shows that the model exhibits a slightly different performance per season when it is applied with the same configuration throughout the year. Hence, if the model is configured according to the season, then it is likely to achieve better performance during its operational implementation. In several case studies, the model has been applied using different configurations from that used in the operational application, giving impressive results, consequently, this knowledge could be useful in this effort. Nevertheless, this procedure has not been applied elsewhere and needs special care, time, and huge computational resources to be accomplished. This effort will also be facilitated by newer model versions including additional features and capabilities. The spatial distribution of statistical scores used was complex and, thus, only a trend of underestimation for western Greece could be partially attributed to the lack of full coupling with dynamically-varying sea processes. As regards the Evrotas and Spercheios rivers, it is important to note that future discharge measurements or estimates could better support a more adequate evaluation of the hydrological forecasts showcasing more specific forecast skill characteristics. Nevertheless, the hydrological response of the system in respect of the water level measurements seems to be sufficient.
The system was also capable of reasonably simulating the severe storm that caused the flash flood of Evrotas river on 7 September 2016, while also simulating high streamflow values reaching ~705 m3 s−1, which caused overflow of a bridge at Skala town. Due to the absence of official flood datasets allowing for detailed validation of critical flood characteristics for the event of 7 September 2016 at the Evrotas river, estimated flood water depth for a specific location was the main criterion used for the verification of the hydraulic–hydrodynamic simulations presented in this work. Thus, in order to examine the model performance, relevant photos and press documentary were used to assess the observed water depth in the location of Skala’s bridge. The river flood modelling results demonstrated in this study show acceptable correspondence between the highest values of the simulated water depth with the empirical evidence obtained after processing information from local authorities and press at Skala’s bridge location.
The integrated modelling system has good potential to be used not only as operational forecasting system but also as a useful research tool in interdisciplinary flood simulation studies. The physically-based modelling system can increase the forecast lead time of weather-driven floods; therefore, it can be exploited as a forecasting guidance for flash flooding management and mitigation design.

Author Contributions

Conceptualization, formal analysis, investigation, methodology, resources, validation, visualization, writing-original draft, writing-review and editing, G.V. and G.P.; conceptualization, formal analysis, investigation, methodology, resources, validation, visualization, vriting-original draft, supervision, writing-review and editing, A.P.; conceptualization, resources, validation, supervision, writing-review and editing, E.D. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Acknowledgments

Several of the calibration simulations of this work have been performed using computational time granted by the Greek Research and Technology Network (GRNET) in the National High-Performance Computer (HPC) facility ARIS (https://hpc.grnet.gr/ (accessed on 12 July 2021)), under project PR007018/HYMETGRID-3. The European Centre for Medium-Range Weather Forecasts (ECMWF) is acknowledged for the provision of 6-h precipitation measurements from 24 Greek land surface meteorological stations included in the network of the World Meteorological Organization (WMO). The National Centers for Environmental Prediction (NCEP) is acknowledged for the provision of the Global Forecasting System (GFS) operational analyses and the real time global (RTG) sea surface temperature (SST) analyses. The National Aeronautics and Space Administration (NASA) and the Hydrological Data and Maps Based on Shuttle Elevation Derivatives at Multiple Scales (HydroSHEDS) are acknowledged for the provision of the Shuttle Radar Topographic Mission (SRTM) digital elevation model (DEM) data. The website wetter3.de (http://www1.wetter3.de/ (accessed on 12 July 2021)) is acknowledged for the provision of the GFS 500-hPa analysis map and the UK Met Office surface analysis map used in this study. The National Observatory of Athens (NOA) is acknowledged for the provision of 24-h precipitation measurements from four Greek land surface meteorological stations. NOA’s 24-h precipitation measurements are freely available from its online archive database (https://meteosearch.meteo.gr/ (accessed on 12 July 2021)). Laconia Live and Laconia’s Municipal Market are also acknowledged for the provision of Evrotas river flood photos. We thank our colleague Leonidas Vardakas for the provision of useful information about Evrotas river. We also thank our colleague Chrysa Vamvakaki for retrieving water level timeseries from the IMBRIW’s archive database and for editing them to be used in this study.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Blöschl, G.; Hall, J.; Viglione, A.; Perdigão, R.A.P.; Parajka, J.; Merz, B.; Lun, D.; Arheimer, B.; Aronica, G.T.; Bilibashi, A.; et al. Changing climate both increases and decreases European river floods. Nature 2019, 573, 108–111. [Google Scholar] [CrossRef]
  2. Berghuijs, W.R.; Aalbers, E.E.; Larsen, J.R.; Trancoso, R.; Woods, R.A. Recent changes in extreme floods across multiple continents. Environ. Res. Lett. 2017, 12, 114035. [Google Scholar] [CrossRef]
  3. Cloke, H.L.; Pappenberger, F. Ensemble flood forecasting: A review. J. Hydrol. 2009, 375, 613–626. [Google Scholar] [CrossRef]
  4. Maidment, D.R. Conceptual Framework for the National Flood Interoperability Experiment. J. Am. Water Resour. Assoc. 2017, 53, 245–257. [Google Scholar] [CrossRef]
  5. Giannaros, C.; Galanaki, E.; Kotroni, V.; Lagouvardos, K.; Oikonomou, C.; Haralambous, H.; Giannaros, T.M. Pre-Operational Application of a WRF-Hydro-Based Fluvial Flood Forecasting System in the Southeast Mediterranean. Forecasting 2021, 3, 26. [Google Scholar] [CrossRef]
  6. Emerton, R.E.; Stephens, E.M.; Pappenberger, F.; Pagano, T.C.; Weerts, A.H.; Wood, A.W.; Salamon, P.; Brown, J.D.; Hjerdt, N.; Donnelly, C.; et al. Continental and global scale flood forecasting systems. Wiley Interdiscip. Rev. Water 2016, 3, 391–418. [Google Scholar] [CrossRef] [Green Version]
  7. Alfieri, L.; Burek, P.; Dutra, E.; Krzeminski, B.; Muraro, D.; Thielen, J.; Pappenberger, F. GloFAS-global ensemble streamflow forecasting and flood early warning. Hydrol. Earth Syst. Sci. 2013, 17, 1161–1175. [Google Scholar] [CrossRef] [Green Version]
  8. Emerton, R.; Zsoter, E.; Arnal, L.; Cloke, H.L.; Muraro, D.; Prudhomme, C.; Stephens, E.M.; Salamon, P.; Pappenberger, F. Developing a global operational seasonal hydro-meteorological forecasting system: GloFAS-Seasonal v1.0. Geosci. Model Dev. 2018, 11, 3327–3346. [Google Scholar] [CrossRef] [Green Version]
  9. Viterbo, F.; Mahoney, K.; Read, L.; Salas, F.; Bates, B.; Elliott, J.; Cosgrove, B.; Dugger, A.; Gochis, D.; Cifelli, R. A multiscale, hydrometeorological forecast evaluation of national water model forecasts of the may 2018 Ellicott City, Maryland, Flood. J. Hydrometeorol. 2020, 21, 475–499. [Google Scholar] [CrossRef]
  10. Gochis, D.; Yu, W.; Yates, D. WRF-Hydro Technical Description and User’s Guide The NCAR WRF-Hydro Technical Description and User’s Guide; NCAR: Boulder, CO, US, 2015. [Google Scholar]
  11. Bartholmes, J.C.; Thielen, J.; Ramos, M.H.; Gentilini, S. Hydrology and Earth System Sciences The European Flood Alert System EFAS-Part 2: Statistical Skill Assessment of Probabilistic and Deterministic Operational Forecasts. Hydrol. Earth Syst. Sci. 2009, 13, 141–153. [Google Scholar] [CrossRef] [Green Version]
  12. Thielen, J.; Bartholmes, J.; Ramos, M.H.; De Roo, A. The European flood alert system–Part 1: Concept and development. Hydrol. Earth Syst. Sci. 2009, 13, 125–140. [Google Scholar] [CrossRef] [Green Version]
  13. Darras, T.; Borrell Estupina, V.; Kong-A-Siou, L.; Vayssade, B.; Johannet, A.; Pistre, S. Identification of spatial and temporal contributions of rainfalls to flash floods using neural network modelling: Case study on the Lez basin (southern France). Hydrol. Earth Syst. Sci. 2015, 19, 4397–4410. [Google Scholar] [CrossRef] [Green Version]
  14. Llasat, M.C.; Marcos, R.; Turco, M.; Gilabert, J.; Llasat-Botija, M. Trends in flash flood events versus convective precipitation in the Mediterranean region: The case of Catalonia. J. Hydrol. 2016, 541, 24–37. [Google Scholar] [CrossRef] [Green Version]
  15. Avolio, E.; Cavalcanti, O.; Furnari, L.; Senatore, A.; Mendicino, G. Brief communication: Preliminary hydro-meteorological analysis of the flash flood of 20 August 2018 in Raganello Gorge, southern Italy. Nat. Hazards Earth Syst. Sci. 2019, 19, 1619–1627. [Google Scholar] [CrossRef] [Green Version]
  16. Petrucci, O.; Aceto, L.; Bianchi, C.; Bigot, V.; Brázdil, R.; Pereira, S.; Kahraman, A.; Kiliç, Ö.; Kotroni, V.; Llasat, M.C.; et al. Flood fatalities in Europe, 1980-2018: Variability, features, and lessons to learn. Water 2019, 11, 1682. [Google Scholar] [CrossRef] [Green Version]
  17. Varlas, G.; Anagnostou, M.N.; Spyrou, C.; Papadopoulos, A.; Kalogiros, J.; Mentzafou, A.; Michaelides, S.; Baltas, E.; Karymbalis, E.; Katsafados, P. A multi-platform hydrometeorological analysis of the flash flood event of 15 November 2017 in Attica, Greece. Remote Sens. 2019, 11, 45. [Google Scholar] [CrossRef] [Green Version]
  18. Papaioannou, G.; Varlas, G.; Papadopoulos, A.; Loukas, A.; Katsafados, P.; Dimitriou, E. Investigating sea-state effects on flash flood hydrograph and inundation forecasting. Hydrol. Process. 2021, 35, e14151. [Google Scholar] [CrossRef]
  19. Diakakis, M.; Deligiannakis, G.; Andreadakis, E.; Katsetsiadou, K.N.; Spyrou, N.I.; Gogou, M.E. How different surrounding environments influence the characteristics of flash flood-mortality: The case of the 2017 extreme flood in Mandra, Greece. J. Flood Risk Manag. 2020, 13, e12613. [Google Scholar] [CrossRef]
  20. Diakakis, M.; Boufidis, N.; Grau, J.; Andreadakis, E.; Stamos, I. A systematic assessment of the effects of extreme flash floods on transportation infrastructure and circulation: The example of the 2017 Mandra flood. Int. J. Disaster Risk Reduct. 2020, 47, 101542. [Google Scholar] [CrossRef]
  21. Diakakis, M.; Andreadakis, E.; Nikolopoulos, E.; Spyrou, N.; Gogou, M.; Deligiannakis, G.; Katsetsiadou, N.; Antoniadis, Z.; Melaki, M.; Georgakopoulos, A.; et al. An integrated approach of ground and aerial observations in flash flood disaster investigations. The case of the 2017 Mandra flash flood in Greece. Int. J. Disaster Risk Reduct. 2019, 33, 290–309. [Google Scholar] [CrossRef]
  22. Spyrou, C.; Varlas, G.; Pappa, A.; Mentzafou, A.; Katsafados, P.; Papadopoulos, A.; Anagnostou, M.N.; Kalogiros, J. Implementation of a nowcasting hydrometeorological system for studying flash flood events: The Case of Mandra, Greece. Remote Sens. 2020, 12, 2784. [Google Scholar] [CrossRef]
  23. Bellos, V.; Papageorgaki, I.; Kourtis, I.; Vangelis, H.; Kalogiros, I.; Tsakiris, G. Reconstruction of a flash flood event using a 2D hydrodynamic model under spatial and temporal variability of storm. Nat. Hazards 2020, 101, 711–726. [Google Scholar] [CrossRef]
  24. Papaioannou, G.; Varlas, G.; Terti, G.; Papadopoulos, A.; Loukas, A.; Panagopoulos, Y.; Dimitriou, E. Flood inundation mapping at ungauged basins using coupled hydrometeorological-hydraulic modelling: The catastrophic case of the 2006 Flash Flood in Volos City, Greece. Water 2019, 11, 2328. [Google Scholar] [CrossRef] [Green Version]
  25. Giannaros, C.; Kotroni, V.; Lagouvardos, K.; Oikonomou, C.; Haralambous, H.; Papagiannaki, K. Hydrometeorological and socio-economic impact assessment of stream flooding in southeast mediterranean: The case of rafina catchment (Attica, Greece). Water 2020, 12, 2426. [Google Scholar] [CrossRef]
  26. Givati, A.; Gochis, D.; Rummler, T.; Kunstmann, H. Comparing One-Way and Two-Way Coupled Hydrometeorological Forecasting Systems for Flood Forecasting in the Mediterranean Region. Hydrology 2016, 3, 19. [Google Scholar] [CrossRef] [Green Version]
  27. Skamarock, W.C.; Klemp, J.B.; Dudhia, J.B.; Gill, D.O.; Barker, D.M.; Duda, M.G.; Huang, X.-Y.; Wang, W.; Powers, J.G. A Description of the Advanced Research WRF Version 3, NCAR Technical Note TN-475+STR; NCAR: Boulder, CO, USA, 2008. [Google Scholar]
  28. US Army Corps of Engineers HEC-RAS River Analysis System; User’s Manual Version 4.1; US Army Corps of Engineers Hydrologic Engineering Center: Davis, CA, USA, 2010; pp. 1–790.
  29. Jiménez, P.A.; Dudhia, J.; González-Rouco, J.F.; Navarro, J.; Montávez, J.P.; García-Bustamante, E. A revised scheme for the WRF surface layer formulation. Mon. Weather Rev. 2012, 140, 898–918. [Google Scholar] [CrossRef] [Green Version]
  30. Hong, S.-Y.; Noh, Y.; Dudhia, J. A new vertical diffusion package with an explicit treatment of entrainment processes. Mon. Weather Rev. 2006, 134, 2318–2341. [Google Scholar] [CrossRef] [Green Version]
  31. Mukul Tewari, N.; Tewari, M.; Chen, F.; Wang, W.; Dudhia, J.; LeMone, M.; Mitchell, K.; Ek, M.; Gayno, G.; Wegiel, J. Implementation and verification of the unified NOAH land surface model in the WRF model (Formerly Paper Number 17.5). In Proceedings of the 20th Conference on Weather Analysis and Forecasting/16th Conference on Numerical Weather Prediction, Seattle, WA, USA, 12–16 January 2004; pp. 11–15. [Google Scholar]
  32. Iacono, M.J.; Delamere, J.S.; Mlawer, E.J.; Shephard, M.W.; Clough, S.A.; Collins, W.D. Radiative forcing by long-lived greenhouse gases: Calculations with the AER radiative transfer models. J. Geophys. Res. Atmos. 2008, 113, 2–9. [Google Scholar] [CrossRef]
  33. Thompson, G.; Field, P.R.; Rasmussen, R.M.; Hall, W.D. Explicit forecasts of winter precipitation using an improved bulk microphysics scheme. Part II: Implementation of a new snow parameterization. Mon. Weather Rev. 2008, 136, 5095–5115. [Google Scholar] [CrossRef]
  34. Grell, G.A.; Freitas, S.R. A scale and aerosol aware stochastic convective parameterization for weather and air quality modeling. Atmos. Chem. Phys. 2014, 14, 5233–5250. [Google Scholar] [CrossRef] [Green Version]
  35. Ek, M.B.; Mitchell, K.E.; Lin, Y.; Rogers, E.; Grunmann, P.; Koren, V.; Gayno, G.; Tarpley, J.D. Implementation of Noah land surface model advances in the National Centers for Environmental Prediction operational mesoscale Eta model. J. Geophys. Res. Atmos. 2003, 108. [Google Scholar] [CrossRef]
  36. Julien, P.Y.; Saghafian, B.; Ogden, F.L. Raster-based hydrologic modeling of spatially-varied surface runoff. JAWRA J. Am. Water Resour. Assoc. 1995, 31, 523–536. [Google Scholar] [CrossRef]
  37. Ogden, F.L. CASC2D Reference Manual; Department of Civil and Environmental Engineering, University of Connecticut: Storrs, CT, USA, 1997; 106p. [Google Scholar]
  38. Garbrecht, J.; Brunner, G. Hydrologic Channel-Flow Routing for Compound Sections. J. Hydraul. Eng. 1991, 117, 629–642. [Google Scholar] [CrossRef]
  39. Jarvis, A.; Guevara, E.; Reuter, H.; Nelson, A. Hole-Filled SRTM for the Globe: Version 4. CGIAR-CSI SRTM 90m Database. Available online: https://research.utwente.nl/en/publications/hole-filled-srtm-for-the-globe-version-4-data-grid (accessed on 12 July 2021).
  40. Lehner, B.; Verdin, K.; Jarvis, A. New global hydrography derived from spaceborne elevation data. Eos 2008, 89, 93–94. [Google Scholar] [CrossRef]
  41. Strahler, A. Hypsometric (area-altitude) analysis of erosional topography. GSA Bull. 1952, 63, 1142. [Google Scholar] [CrossRef]
  42. Camera, C.; Bruggeman, A.; Zittis, G.; Sofokleous, I.; Arnault, J. Simulation of extreme rainfall and streamflow events in small Mediterranean watersheds with a one-way-coupled atmospheric-hydrologic modelling system. Nat. Hazards Earth Syst. Sci. 2020, 20, 2791–2810. [Google Scholar] [CrossRef]
  43. Senatore, A.; Mendicino, G.; Gochis, D.J.; Yu, W.; Yates, D.N.; Kunstmann, H. Fully coupled atmosphere-hydrology simulations for the central Mediterranean: Impact of enhanced hydrological parameterization for short and long time scales. J. Adv. Model. Earth Syst. 2015, 7, 1693–1715. [Google Scholar] [CrossRef]
  44. Senatore, A.; Furnari, L.; Mendicino, G. Impact of high-resolution sea surface temperature representation on the forecast of small Mediterranean catchments’ hydrological responses to heavy precipitation. Hydrol. Earth Syst. Sci. 2020, 24, 269–291. [Google Scholar] [CrossRef] [Green Version]
  45. Yucel, I.; Onen, A.; Yilmaz, K.K.; Gochis, D.J. Calibration and evaluation of a flood forecasting system: Utility of numerical weather prediction model, data assimilation and satellite-based rainfall. J. Hydrol. 2015, 523, 49–66. [Google Scholar] [CrossRef] [Green Version]
  46. Ozkaya, A.; Akyurek, Z. WRF-Hydro Model Application in a Data-Scarce, Small and Topographically Steep Catchment in Samsun, Turkey. Arab. J. Sci. Eng. 2020, 45, 3781–3798. [Google Scholar] [CrossRef]
  47. Brunner, G.W. HEC-RAS River Analysis System, 2D Modeling User’s Manual Version 5.0; US Army Corps of Engineers–Hydrologic Engineering Center: Davis, CA, USA, 2016; pp. 1–171.
  48. Muthusamy, M.; Casado, M.R.; Butler, D.; Leinster, P. Understanding the effects of Digital Elevation Model resolution in urban fluvial flood modelling. J. Hydrol. 2021, 596. [Google Scholar] [CrossRef]
  49. Azeez, O.; Elfeki, A.; Kamis, A.S.; Chaabani, A. Dam break analysis and flood disaster simulation in arid urban environment: The Um Al-Khair dam case study, Jeddah, Saudi Arabia. Nat. Hazards 2020, 100, 995–1011. [Google Scholar] [CrossRef]
  50. Khanam, M.; Sofia, G.; Koukoula, M.; Lazin, R.; Nikolopoulos, E.I.; Shen, X.; Anagnostou, E.N. Impact of compound flood event on coastal critical infrastructures considering current and future climate. Nat. Hazards Earth Syst. Sci. 2021, 21, 587–605. [Google Scholar] [CrossRef]
  51. Urzică, A.; Mihu-Pintilie, A.; Stoleriu, C.C.; Cîmpianu, C.I.; Huţanu, E.; Pricop, C.I.; Grozavu, A. Using 2D HEC-RAS modeling and embankment dam break scenario for assessing the flood control capacity of a multireservoir system (Ne Romania). Water 2021, 13, 57. [Google Scholar] [CrossRef]
  52. Psomiadis, E.; Tomanis, L.; Kavvadias, A.; Soulis, K.X.; Charizopoulos, N.; Michas, S. Potential dam breach analysis and flood wave risk assessment using HEC-RAS and remote sensing data: A multicriteria approach. Water 2021, 13, 364. [Google Scholar] [CrossRef]
  53. Papaioannou, G.; Vasiliades, L.; Loukas, A.; Alamanos, A.; Efstratiadis, A.; Koukouvinos, A.; Tsoukalas, I.; Kossieris, P. A flood inundation modeling approach for urban and rural areas in lake and large-scale river basins. Water 2021, 13, 1264. [Google Scholar] [CrossRef]
  54. Papaioannou, G.; Efstratiadis, A.; Vasiliades, L.; Loukas, A.; Papalexiou, S.M.; Koukouvinos, A.; Tsoukalas, I.; Kossieris, P. An operational method for Flood Directive implementation in ungauged urban areas. Hydrology 2018, 5, 24. [Google Scholar] [CrossRef] [Green Version]
  55. Brunner, G.W.; Sanchez, A.; Molls, T.; Parr, D.A. HEC-RAS Verification and Validation Tests; US Army Corps of Engineers–Hydrologic Engineering Center: Davis, CA, USA, 2018; pp. 1–154.
  56. Brunner, G.W. CEIWR-HHT Benchmarking of the HEC-RAS Two-Dimensional Hydraulic Modeling Capabilities; US Army Corps of Engineers–Hydrologic Engineering Center: Davis, CA, USA, 2018.
  57. Papaioannou, G.; Loukas, A.; Vasiliades, L.; Aronica, G.T. Flood inundation mapping sensitivity to riverine spatial resolution and modelling approach. Nat. Hazards 2016, 83, 117–132. [Google Scholar] [CrossRef]
  58. Schumann, G.; Matgen, P.; Cutler, M.E.J.; Black, A.; Hoffmann, L.; Pfister, L. Comparison of remotely sensed water stages from LiDAR, topographic contours and SRTM. ISPRS J. Photogramm. Remote Sens. 2008, 63, 283–296. [Google Scholar] [CrossRef]
  59. Wang, W.; Yang, X.; Yao, T. Evaluation of ASTER GDEM and SRTM and their suitability in hydraulic modelling of a glacial lake outburst flood in southeast Tibet. Hydrol. Process. 2012, 26, 213–225. [Google Scholar] [CrossRef]
  60. Yalcin, E. Assessing the impact of topography and land cover data resolutions on two-dimensional HEC-RAS hydrodynamic model simulations for urban flood hazard analysis. Nat. Hazards 2020, 101, 995–1017. [Google Scholar] [CrossRef]
  61. Lim, N.J.; Brandt, S.A. Flood map boundary sensitivity due to combined effects of DEM resolution and roughness in relation to model performance. Geomatics Nat. Hazards Risk 2019, 10, 1613–1647. [Google Scholar] [CrossRef] [Green Version]
  62. Brunner, G. HEC-RAS River Analysis System, Hydraulic Reference Manual, Version 4.1; US Army Corps of Engineers–Hydrologic Engineering Center: Davis, CA, USA, 2010; pp. 1–790.
  63. Warner, J.C.; Brunner, G.W.; Wolfe, B.C.; Piper, S.S. HEC-RAS. River Analysis System. Applications Guide. Version 4.0; US Army Corps of Engineers–Hydrologic Engineering Center: Davis, CA, USA, 2008.
  64. Water, S.S. Flood Risk Management Plans of Eastern Peloponnese River Basin District. Flood Risk Maps. ANNEX 8—Topographic Surveys of Technical Works—Sections; Ministry of Environment and Energy (SSW-MEE): Athens, Greece, 2018.
  65. Laconia Live The Skala’s Bridge “Disappeared”. 2016. Available online: https://laconialive.gr/εξαφανίστηκε-η-γέφυρα-της-σκάλας/ (accessed on 12 July 2021). (In Greek).
  66. Laconia’s Municipal Market Newspaper Municipality of Eurota “FLOOD 2016”. 2016. Available online: http://www.dimotikiagoratislakonias.gr/epikairotita/aftodioikisi/dimoi/item/1018-dimos-evrota-plimmyra-2016 (accessed on 12 July 2021). (In Greek).
  67. Papadopoulos, A.; Katsafados, P. Verification of operational weather forecasts from the POSEIDON system across the Eastern Mediterranean. Nat. Hazards Earth Syst. Sci. 2009, 9, 1299–1306. [Google Scholar] [CrossRef] [Green Version]
  68. Soukissian, T.; Papadopoulos, A. Effects of different wind data sources in offshore wind power assessment. Renew. Energy 2015, 77, 101–114. [Google Scholar] [CrossRef]
  69. Wilks, D. Statistical Methods in the Atmospheric Sciences; Academic Press: Oxford, UK, 2011; Volume 100, p. 704. [Google Scholar]
  70. Papadopoulos, A.; Chronis, T.G.; Anagnostou, E.N. Improving convective precipitation forecasting through assimilation of regional lightning measurements in a mesoscale model. Mon. Weather Rev. 2005, 133, 1961–1977. [Google Scholar] [CrossRef] [Green Version]
  71. Mentzafou, A.; Varlas, G.; Papadopoulos, A.; Poulis, G.; Dimitriou, E. Assessment of Automatically Monitored Water Levels and Water Quality Indicators in Rivers with Different Hydromorphological Conditions and Pollution Levels in Greece. Hydrology 2021, 8, 86. [Google Scholar] [CrossRef]
  72. Nash, J.E.; Sutcliffe, J.V. River flow forecasting through conceptual models part I—A discussion of principles. J. Hydrol. 1970, 10, 282–290. [Google Scholar] [CrossRef]
  73. Katsafados, P.; Papadopoulos, A.; Mavromatidis, E.; Gikas, N. Quantitative Verification Statistics of WRF Predictions over the Mediterranean Region. Available online: http://www.erasmus.hua.gr/~meteoclima/images/stories/model/12WRF2011.extabs.pdf (accessed on 12 July 2021).
  74. Snook, N.; Kong, F.; Brewster, K.A.; Xue, M.; Thomas, K.W.; Supinie, T.A.; Perfater, S.; Albright, B. Evaluation of convection-permitting precipitation forecast products using WRF, NMMB, and FV3 for the 2016-17 NOAA hydrometeorology testbed flash flood and intense rainfall experiments. Weather Forecast. 2019, 34, 781–804. [Google Scholar] [CrossRef] [Green Version]
  75. Katsafados, P.; Papadopoulos, A.; Korres, G.; Varlas, G. A fully coupled atmosphere-ocean wave modeling system for the Mediterranean Sea: Interactions and sensitivity to the resolved scales and mechanisms. Geosci. Model Dev. 2016, 9, 161–173. [Google Scholar] [CrossRef] [Green Version]
  76. Katsafados, P.; Varlas, G.; Papadopoulos, A.; Spyrou, C.; Korres, G. Assessing the Implicit Rain Impact on Sea State During Hurricane Sandy (2012). Geophys. Res. Lett. 2018, 45, 12015–12022. [Google Scholar] [CrossRef]
  77. Varlas, G.; Katsafados, P.; Papadopoulos, A.; Korres, G. Implementation of a two-way coupled atmosphere-ocean wave modeling system for assessing air-sea interaction over the Mediterranean Sea. Atmos. Res. 2018, 208, 201–217. [Google Scholar] [CrossRef]
  78. Varlas, G.; Spyrou, C.; Papadopoulos, A.; Korres, G.; Katsafados, P. One-year assessment of the CHAOS two-way coupled atmosphere-ocean wave modelling system over the Mediterranean and Black Seas. Mediterr. Mar. Sci. 2020, 21, 372–385. [Google Scholar] [CrossRef]
  79. Varlas, G.; Vervatis, V.; Spyrou, C.; Papadopoulou, E.; Papadopoulos, A.; Katsafados, P. Investigating the impact of atmosphere–wave–ocean interactions on a Mediterranean tropical-like cyclone. Ocean Model. 2020, 153, 101675. [Google Scholar] [CrossRef]
  80. Varlas, G.; Marinou, E.; Gialitaki, A.; Siomos, N.; Tsarpalis, K.; Kalivitis, N.; Solomos, S.; Tsekeri, A.; Spyrou, C.; Tsichla, M.; et al. Assessing sea-state effects on sea-salt aerosol modeling in the lower atmosphere using lidar and in-situ measurements. Remote Sens. 2021, 13, 614. [Google Scholar] [CrossRef]
Atmosphere 12 00902 g001 550
Figure 1. Flowchart of main data used and model setup in IMBRIW’s hydrometeorological forecasting system. Models, parameters, and data are described in the following sub-sections.
Figure 1. Flowchart of main data used and model setup in IMBRIW’s hydrometeorological forecasting system. Models, parameters, and data are described in the following sub-sections.
Atmosphere 12 00902 g001
Atmosphere 12 00902 g002 550
Figure 2. (a) Meteorological model domains (D1: 36 km × 36 km, D2: 12 km × 12 km, D3: 4 km × 4 km, Spercheios: 1 km × 1 km, and Evrotas: 1 km × 1 km); Topography (m) in the grid spacing of 100 m × 100 m and stream order used by the hydrological model for the simulations in (b) Evrotas and (c) Spercheios river basins. Sentenikos and Anthili areas where IMBRIW operates hydrological stations are also shown. A bridge located at Skala town, which is crossed by the Evrotas river, is also depicted. Moreover, only stream orders from 2 to 6 are illustrated.
Figure 2. (a) Meteorological model domains (D1: 36 km × 36 km, D2: 12 km × 12 km, D3: 4 km × 4 km, Spercheios: 1 km × 1 km, and Evrotas: 1 km × 1 km); Topography (m) in the grid spacing of 100 m × 100 m and stream order used by the hydrological model for the simulations in (b) Evrotas and (c) Spercheios river basins. Sentenikos and Anthili areas where IMBRIW operates hydrological stations are also shown. A bridge located at Skala town, which is crossed by the Evrotas river, is also depicted. Moreover, only stream orders from 2 to 6 are illustrated.
Atmosphere 12 00902 g002
Atmosphere 12 00902 g003 550
Figure 3. Hydraulic modelling domain at Skala’s bridge (a) Hydraulic modelling application study area and the effect of bridge design in the mesh generation, (b) UAV aerial photographs of the study area, (c) Topographical survey sketch [64], (d) Technical details of the bridge piers and abutments, (e) Topographical survey photographs [64], and (f) Photographs of Skala’s bridge during the flash flood event studied [65,66].
Figure 3. Hydraulic modelling domain at Skala’s bridge (a) Hydraulic modelling application study area and the effect of bridge design in the mesh generation, (b) UAV aerial photographs of the study area, (c) Topographical survey sketch [64], (d) Technical details of the bridge piers and abutments, (e) Topographical survey photographs [64], and (f) Photographs of Skala’s bridge during the flash flood event studied [65,66].
Atmosphere 12 00902 g003
Atmosphere 12 00902 g004 550
Figure 4. Map presenting the locations of 24 land surface meteorological stations used in the evaluation of precipitation forecasts. The meteorological stations are included in the network of the World Meteorological Organization and the precipitation measurements were provided by the European Centre for Medium-Range Weather Forecasts. Topography (m) of the D3 (4 km × 4 km) meteorological domain is also illustrated.
Figure 4. Map presenting the locations of 24 land surface meteorological stations used in the evaluation of precipitation forecasts. The meteorological stations are included in the network of the World Meteorological Organization and the precipitation measurements were provided by the European Centre for Medium-Range Weather Forecasts. Topography (m) of the D3 (4 km × 4 km) meteorological domain is also illustrated.
Atmosphere 12 00902 g004
Atmosphere 12 00902 g005 550
Figure 5. Precipitation (mm) forecast skill indicated by the dependence of (a) BIAS and (b) ETS to forecast horizon (6–120 h). Only 6-h accumulated precipitation values over 1 mm were considered in the estimation of the contingency table.
Figure 5. Precipitation (mm) forecast skill indicated by the dependence of (a) BIAS and (b) ETS to forecast horizon (6–120 h). Only 6-h accumulated precipitation values over 1 mm were considered in the estimation of the contingency table.
Atmosphere 12 00902 g005
Atmosphere 12 00902 g006 550
Figure 6. Diagrams of (a) BIAS and (b) ETS for 6-h accumulated precipitation (mm) over all stations for the first forecast day (from the 12th to the 36th forecast hour). Y-axis represents BIAS and ETS, respectively, while X-axis represents 6-h accumulated precipitation (mm) thresholds considered in the estimation of the contingency table.
Figure 6. Diagrams of (a) BIAS and (b) ETS for 6-h accumulated precipitation (mm) over all stations for the first forecast day (from the 12th to the 36th forecast hour). Y-axis represents BIAS and ETS, respectively, while X-axis represents 6-h accumulated precipitation (mm) thresholds considered in the estimation of the contingency table.
Atmosphere 12 00902 g006
Atmosphere 12 00902 g007 550
Figure 7. Diagrams of BIAS for 6-h accumulated precipitation (mm) over all stations on (a) winter, (b) spring, (c) summer, and (d) autumn; (e–h) the same for ETS. Y-axis represents BIAS and ETS, respectively, while X-axis represents 6-h accumulated precipitation (mm) thresholds considered in the estimation of the contingency table.
Figure 7. Diagrams of BIAS for 6-h accumulated precipitation (mm) over all stations on (a) winter, (b) spring, (c) summer, and (d) autumn; (e–h) the same for ETS. Y-axis represents BIAS and ETS, respectively, while X-axis represents 6-h accumulated precipitation (mm) thresholds considered in the estimation of the contingency table.
Atmosphere 12 00902 g007
Atmosphere 12 00902 g008 550
Figure 8. Map with all stations colored to demonstrate 6-h accumulated precipitation (a) BIAS and (b) ETS. Only 6-h accumulated precipitation (mm) values over 1 mm were considered in the estimation of the contingency table for each station.
Figure 8. Map with all stations colored to demonstrate 6-h accumulated precipitation (a) BIAS and (b) ETS. Only 6-h accumulated precipitation (mm) values over 1 mm were considered in the estimation of the contingency table for each station.
Atmosphere 12 00902 g008
Atmosphere 12 00902 g009 550
Figure 9. Timeseries of daily average water level (cm) at (a) the Sentenikos station (Evrotas river) and (b) the Anthili station (Spercheios river). Simulated and measured values are represented by blue and black colors, respectively. Y-axis represents water level (cm) while x-axis represents time (month/year). MBE, RMSE, and NSE statistical scores are shown at the top-left corner of each plot.
Figure 9. Timeseries of daily average water level (cm) at (a) the Sentenikos station (Evrotas river) and (b) the Anthili station (Spercheios river). Simulated and measured values are represented by blue and black colors, respectively. Y-axis represents water level (cm) while x-axis represents time (month/year). MBE, RMSE, and NSE statistical scores are shown at the top-left corner of each plot.
Atmosphere 12 00902 g009
Atmosphere 12 00902 g010 550
Figure 10. (a) Geopotential height (black contours in gpm) and temperature (color-shaded areas and white contours in °C) at the isobaric level of 500 hPa on 7 September 2016 at 06:00 UTC. The map is based on analysis data from the Global Forecast System (GFS); (b) UK Met Office surface analysis map (mean sea level pressure in hPa) on 7 September 2016 at 06:00 UTC. The maps were derived from the archive of wetter3.de (http://www1.wetter3.de/ (accessed on 12 July 2021)).
Figure 10. (a) Geopotential height (black contours in gpm) and temperature (color-shaded areas and white contours in °C) at the isobaric level of 500 hPa on 7 September 2016 at 06:00 UTC. The map is based on analysis data from the Global Forecast System (GFS); (b) UK Met Office surface analysis map (mean sea level pressure in hPa) on 7 September 2016 at 06:00 UTC. The maps were derived from the archive of wetter3.de (http://www1.wetter3.de/ (accessed on 12 July 2021)).
Atmosphere 12 00902 g010
Atmosphere 12 00902 g011 550
Figure 11. Spatial distribution of 24-h accumulated precipitation (mm) resulting from the meteorological forecast for the period from 6 September at 21:00 to 7 September at 21:00 UTC (corresponding to the period from 7 September at 00:00 to 8 September at 00:00 local time, i.e., UTC + 3 h). The map also shows precipitation measurements from land surface meteorological stations at Geraki, Sparta, Molaoi, and Kardamyli for the same period. These stations are included in the network of the National Observatory of Athens (NOA) and their measurements are available at the NOA’s archive database (https://meteosearch.meteo.gr/ (accessed on 12 July 2021)).
Figure 11. Spatial distribution of 24-h accumulated precipitation (mm) resulting from the meteorological forecast for the period from 6 September at 21:00 to 7 September at 21:00 UTC (corresponding to the period from 7 September at 00:00 to 8 September at 00:00 local time, i.e., UTC + 3 h). The map also shows precipitation measurements from land surface meteorological stations at Geraki, Sparta, Molaoi, and Kardamyli for the same period. These stations are included in the network of the National Observatory of Athens (NOA) and their measurements are available at the NOA’s archive database (https://meteosearch.meteo.gr/ (accessed on 12 July 2021)).
Atmosphere 12 00902 g011
Atmosphere 12 00902 g012 550
Figure 12. Spatial distribution of 1-h accumulated precipitation (mm) resulting from the meteorological forecast and Evrotas river streamflow (m3 s−1) resulting from the hydrological forecast for 7 September at (a) 02:00, (b) 03:00, (c) 04:00, (d) 05:00, (e) 06:00, (f) 07:00, (g) 08:00, (h) 09:00, and (i) 10:00 UTC.
Figure 12. Spatial distribution of 1-h accumulated precipitation (mm) resulting from the meteorological forecast and Evrotas river streamflow (m3 s−1) resulting from the hydrological forecast for 7 September at (a) 02:00, (b) 03:00, (c) 04:00, (d) 05:00, (e) 06:00, (f) 07:00, (g) 08:00, (h) 09:00, and (i) 10:00 UTC.
Atmosphere 12 00902 g012
Atmosphere 12 00902 g013 550
Figure 13. Inundated areas and overflow on Skala’s bridge as simulated by the HEC-RAS hydraulic–hydrodynamic model on 7 September 2016 at 08:00 UTC. (a) Spatial distribution of the maximum water depth, (b) the maximum water depth at the profile line (a)—green color line), (c) time series of maximum water depth at the deepest part of the profile line (a)—green color line).
Figure 13. Inundated areas and overflow on Skala’s bridge as simulated by the HEC-RAS hydraulic–hydrodynamic model on 7 September 2016 at 08:00 UTC. (a) Spatial distribution of the maximum water depth, (b) the maximum water depth at the profile line (a)—green color line), (c) time series of maximum water depth at the deepest part of the profile line (a)—green color line).
Atmosphere 12 00902 g013
Table
Table 1. Manning roughness coefficient (Manning, dimensionless), channel bottom width (CBW in m), and slide slope (CSS, dimensionless) of Evrotas and Spercheios channels dependent on stream order classification.
Table 1. Manning roughness coefficient (Manning, dimensionless), channel bottom width (CBW in m), and slide slope (CSS, dimensionless) of Evrotas and Spercheios channels dependent on stream order classification.
Stream OrderManningCBW (m)CSS
10.6521.0
20.550.6
30.4100.3
40.3200.18
50.1300.05
60.03500.05
Table
Table 2. Information (WMO ID, latitude, longitude, and altitude) for the land surface meteorological stations used in the evaluation of precipitation forecasts.
Table 2. Information (WMO ID, latitude, longitude, and altitude) for the land surface meteorological stations used in the evaluation of precipitation forecasts.
WMO IDLatitude (°N)Longitude (°E)Altitude (m)WMO IDLatitude (°N)Longitude (°E)Altitude (m)
1662240.5222.9741670638.3326.134
1662440.9824.651671538.123.78235
1662740.8525.9231671838.0723.5531
1664139.6219.9241672337.726.927
1664239.720.824831672637.0722.028
1664338.6220.7741674137.9223.9272
1664839.6322.42741674236.7827.07129
1665039.9225.2341674635.4824.12151
1666739.0726.651674936.428.0811
1668237.9221.28141675435.3325.1839
1668438.9724.48281676035.1825.32336
1668538.1220.5221676535.5227.2520
Table
Table 3. Contingency table. The table elements are defined as follows. A: model forecast and measurement exceeded the threshold; B: model forecast exceeded the threshold but measurement did not; C: model forecast did not reach the threshold but measurement exceeded it; D: model forecast and measurement did not reach the threshold.
Table 3. Contingency table. The table elements are defined as follows. A: model forecast and measurement exceeded the threshold; B: model forecast exceeded the threshold but measurement did not; C: model forecast did not reach the threshold but measurement exceeded it; D: model forecast and measurement did not reach the threshold.
Event ForecastEvent Observed
YesNo
YesAB
NoCD
Table
Table 4. Streamflow (m3 s−1) on 7 September 2016 per hour as simulated by the WRF-Hydro hydrological model for the river location (36.8602° N, 22.6677° E) used to estimate the input hydrograph of the HEC-RAS hydraulic–hydrodynamic simulation.
Table 4. Streamflow (m3 s−1) on 7 September 2016 per hour as simulated by the WRF-Hydro hydrological model for the river location (36.8602° N, 22.6677° E) used to estimate the input hydrograph of the HEC-RAS hydraulic–hydrodynamic simulation.
Hour (UTC)Streamflow (m3 s−1)Hour (UTC)Streamflow (m3 s−1)Hour (UTC)Streamflow (m3 s−1)
000.508704.6616250.74
010.5609704.9617223.49
020.5910596.6518196.72
030.6711555.419170.8
041.412485.6320147.72
051.6913401.0621128.14
062.3314339.4622111.81
072.3415288.382398.46
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Varlas, G.; Papadopoulos, A.; Papaioannou, G.; Dimitriou, E. Evaluating the Forecast Skill of a Hydrometeorological Modelling System in Greece. Atmosphere 2021, 12, 902. https://doi.org/10.3390/atmos12070902

AMA Style

Varlas G, Papadopoulos A, Papaioannou G, Dimitriou E. Evaluating the Forecast Skill of a Hydrometeorological Modelling System in Greece. Atmosphere. 2021; 12(7):902. https://doi.org/10.3390/atmos12070902

Chicago/Turabian Style

Varlas, George, Anastasios Papadopoulos, George Papaioannou, and Elias Dimitriou. 2021. "Evaluating the Forecast Skill of a Hydrometeorological Modelling System in Greece" Atmosphere 12, no. 7: 902. https://doi.org/10.3390/atmos12070902

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop