Introduction

Hydrologic and hydraulic calculations are essential for the practical applications of the hydraulic design of flood protection structures. The design rainfall is a required input for the hydrologic and hydraulic design calculations. The design rainfall is distributed over the storm duration (hyetograph) by various rainfall distribution patterns (RDPs) to generate the runoff hydrographs. The runoff hydrographs are further routed in the watershed from upstream to downstream through the hydraulic modeling for producing flood hydrographs at various locations along the main stream. The resulting flood hydrographs are used for investigating the response of channels, streams, and other infrastructure in the watershed for different storm events. This establishes the fact that the suitable design rainfall pattern plays an important role in the hydrologic and hydraulic analysis.

Different distributions have been tested for different climates of the world. Widely used empirical RDPs were developed by Natural Resources Conservation Service (NRCS). Such RDPs are called the U.S. Soil Conservation Service (SCS) RDPs. The SCS developed charts for estimating instantaneous peak discharges resulting from the small watershed. The peak discharge rates were used for analysis of water control measures. SCS method presents four different types of 24-h RDPs, namely, SCS types I, IA, II, and III (SCS 1986). Typically, SCS types I and IA are recommended for Alaska, Hawaii, and the coastal side of the Sierra Nevada and Cascade mountainous region, while for the rest of the USA, Puerto Rico, and the Virgin Islands, the SCS type II RDPs are recommended. The use of SCS type III is recommended for the hurricane-affected areas (Guo and Hargadin 2009).

The use of SCS-type RDPs to other parts of the world is questionable. Yet, most practitioners use SCS type RDPs in their studies even though it may not be applicable in their study area. The alternative way is to use the design RDPs. The derivation of such distributions is generally based on the precipitation depth–distribution–frequency (DDF) or intensity–distribution–frequency (IDF) relationships that describe the burst with the most intensity during different rainfall durations (Pilgrim and Cordery 1975; Ewea et al. 2017). The SCS-type RDPs are used to simulate DDF or IDF curves for the 24-h duration. Huff (1967) used the time-distribution characteristics of heavy storms to develop the families of curves for designing the RDPs for approximately 400 mi2 areas in Illinois and the Midwest. The design RDPs were recommended for evaluating the design and operation of runoff control structures and assessing the storm events in weather modification operations. Different procedures are available for constructing the design RDPs. The most notably used procedures are constructing the RDPs based on the DDF relationship, suggested by the Hershfield (1961); constructing the design distributions based on the time quartile technique using the observed storm events, presented by Huff (1967); and constructing the distributions based on the hyetograph method, explained by Kerr et al. (1974). Apart from these procedures, Guo and Hargadin (2009) suggested another technique for designing the rainfall distributions by using the concept of enveloping curves which is based on the similar approach as was used for constructing SCS curves; Humphreys et al. (2013) used rainfall intensity formulas and a log–log distribution for combining 6-h and 24-h rainfall into a single storm distribution hydrograph in order to utilize a hypothetical nested rainfall distribution for the hydrologic computations; Dullo et al. (2017) proposed a framework for constructing the site-specific temporal RDPs using a clustering technique.

Rainfall processes in arid regions, especially in most parts of Saudi Arabia, are characterized by large random special and temporal distributions, and further influenced by a climate change. Recently, climate has been cited as one of the significant challenges for flood risk analysis and water supply works (Ishak et al. 2013). The analyses of different rainfall events present dramatic changes with respect to the rainfall events that occur in humid regions. Most of the existing methodologies for humid regions usually require modifications in order to be applicable to arid regions (Ewea et al. 2016). The use of the SCS-type RDPs is further complicated by the data coverage and length limitations in arid regions. Runoff in such regions is generated from random localized rainfall events, exhibiting high intensities with short durations. The most common of them is the convective type rainfall events, resulting from unstable synoptic conditions (Yair and Lavee 1985; El-Hames and Richards 1998). The generated hydrograph from the convective type rainfall is characterized by a steep rising limb with a short time to peak, and a rapid recession to zero base flow. The major hydrological factors that influence the runoff occurrences are high-intensity rainfall, geological and geomorphological features, initial abstractions, and transmission losses in the dry channel (Gheith and Sultan 2002).

The suitable RDPs play an important role in the analysis and design of hydrology, hydraulics, and water resources projects. Especially, it is important for the flood risk analysis. The flooding phenomenon cannot be ignored because it is established that different parts of the world have been experiencing increased flooding risks from extreme rainfall events (Kusratmoko et al. 2016). Floods usually resulted in the loss of human lives and in extensive destructions of infrastructures, especially in urban areas. At the global level, more than one third of the urban areas are where 82% of the population resides in the flood-prone areas (Dilley et al. 2005). It is known that occurrences of extreme rainfall-flooding events have been associated with an extreme weather phenomenon that is common in arid regions. Flooding in the form of flash floods is influenced by the random rainfall events in arid regime characteristics, exhibiting large variations in time and space, with high intensities and short durations. The generating flood flows have high velocities and high peak discharge, with small time to peak. The characteristic of small time to peak contributes to the severe destruction of infrastructures from flooding events, which commonly happens in the arid regions (Elfeki and Bahrawi 2017).

Hydrologic and hydraulic modeling are usually carried out using software. Eslami et al. (2017) used the Hydrologic Engineering Center’s Hydrologic Modeling System (HEC-HMS version 4.2.1) to estimate the flood hydrographs for nine sub-basins of Eskandari catchment in Iran. They investigated the prioritization of the sub-basins in terms of flood risk. Cheshmidari et al. (2017) employed Delphi technique to investigate the intensity of flood in Ivar watershed in Iran. They concluded that the technique is a useful planning tool to manage flood hazards. Bahremand et al. (2017) presented an adaptive neuro-fuzzy interface system to model flood hydrograph of Shirindarreh river. They concluded that the presented model gave reasonable results; however, lack of data of other parameters such as land use and rainfall could improve the results. Beretta et al. (2018) used two-dimensional model to simulate a real flood and compared the results to a physical laboratory model. Several methods to simulate building were used. They found out that an adjusted roughness coefficient gave acceptable results when compared with the detailed geometry. Rebolho et al. (2018) presented a simple model to produce inundation maps for data-scarce areas. The model is based on Digital Elevation Model (DEM) analysis and one-dimensional hydraulic equations. However, they stated that this approach could be used for rapid production of inundation maps but cannot replace more detailed methods.

This research paper investigates the importance of temporal RDPs in arid regions based on a case study on Taibah and Islamic Universities (TU-IU) catchment (Abdulrazzak et al. 2018). The construction of the design temporal RDPs is very important in the arid regions because the post rainfall-runoff processes such as the flood analysis depend upon the suitable rainfall distribution patterns in the study area. The conventional method in the rainfall-runoff analysis is based on SCS type II method, which is compared with the design temporal RDPs in this study. The results advocate that the use of SCS type II has no justification in the arid regions and it is associated with significant overestimation of peak discharges. The paper addresses all the implications arising from the use of RDPs by further investigating the flood inundation parameters (peak discharge, time to peak, flood volume, inundation depth and width, and flow velocity). Comparison of flood inundation parameters resulting from the simulations of the two different types of RDPs is analyzed further to establish the importance and significance of the suitable RDPs in arid regions.

Study area and rainfall characteristics

The study area is a small catchment covering TU-IU campuses in Medina region, Saudi Arabia as shown in Fig. 1. The catchment is located at N 24° 28.788′ and E 39° 32.767′, and its topography is characterized by the gentle elevation surrounded by relatively high mountains in the upstream area and flat area with some hills covering both universities (Figs. 1 and 2). The ground elevation varies from 533 to 966 m. The topographic characteristics indicate that the catchment has an average slope of 1%, orienting from west to east direction toward the downstream of the main channel (wadi Al Aqiq). The geological characteristics of the catchment revealed that the geology mainly consists of a barren surface with hard igneous and metamorphic rocks, limited exposure of the surface rocks in some upstream areas, and the barren vegetation. The catchment bounds the channel that is flowing along the Taibah University and the Islamic University. The channel is a composite channel, comprising a natural earthen channel, where the Taibah University is located to the North of the channel, and the artificial man-made concrete channel, where the Islamic University is located to the South of the channel. The same TU-IU channel was analyzed and modeled for flood inundation mapping. Both universities are surrounded by the urban infrastructure like buildings and roads, as shown in Fig. 2.

Fig. 1
figure 1

Location of the study area

Full size image
Fig. 2
figure 2

Infrastructures in and around the TU-IU and the flood channel from satellite imagery

Full size image

Rainfall characteristics were estimated using the records of two main stations M001 and M103 as shown in Fig. 3. Station M001 is located in Sultanah district, which is 4.3 km northeast from the TU-IU catchment (with 48 years of records), while station M103 is located in Bir-Almashi. M001 station was established in 1968 (with 43 years of records), with its elevation at 590 m above the mean sea level and average annual rainfall at 60 mm. The rainfall events mainly occur in spring season (March, April, and May), as shown in Fig. 4, with an average seasonal depth of 35 mm. Further evaluation of rainfall characteristics revealed the random nature of rainfall events, with their occurrences mainly in the months of January, April, March, and November. The high values were recorded in months of March and April. Average rainfall depths ranged from 18 to 29 mm.

Fig. 3
figure 3

Location of rainfall stations M001 and M103

Full size image
Fig. 4
figure 4

Average monthly rainfall depth for station M001

Full size image

Methodology

Figure 5 shows the methodology used in the current study. The data consisted of the rainfall records from the two rainfall gauges (M001 and M103). Different types of statistical distributions were used for performing the frequency analysis. Various distributions were compared in order to select the best probability distribution function based on the root mean square error (RMSE) criterion. The inverse distance weighing (IDW) interpolation technique was used to distribute the maximum rainfall depths from the two rainfall stations over the whole catchment for different return periods. The next important task was to develop the design RDPs for the study region. The designed RDPs provide curves for estimating the hyetographs for various return periods. Other characteristics of the catchment were found from the watershed modeling of the TU-IU catchment using Watershed Modeling System (WMS) software. In the succeeding steps, hydrological and subsequent hydraulic modelings were performed. The impacts of the different RDPs on hydrologic modeling and hydraulic modeling were analyzed in the last step. A comparison between the results obtained from SCS type II RDPs and the designed RDPs was made. The results include the hydrological and hydraulic modeling parameters, namely, time to peak, peak discharge, flood volume, top inundation width, inundation depth, and flow velocity.

Fig. 5
figure 5

Schematic working flow of the used methodology

Full size image

Frequency analysis

Frequency analysis was made for the two rainfall records (i.e., M001 and M103) having more than 30 years of data. Different statistical distributions were applied: normal distribution, Gumbel distribution type 1, generalized extreme value (GEV) distribution, two-parameter and three-parameter log-normal distributions, Pearson type III distribution, and log-Pearson type III distributions. The most suitable distribution that fits the rainfall record relatively better among others was chosen for the study area using Storm Management And Design Aid (SMADA) program. The comparison is based on RMSE statistical technique in which the distribution method that resulted in the least RMSE value was chosen as the relatively best distribution method (Kite 1977),

$$ RMSE=\sqrt{\frac{1}{n}{\sum}_{i=1}^n{\left[{\overline{R}}_i-{R}_i\right]}^2} $$
(1)

where Ri is the observed rainfall depth (Ministry of Environment, Water and Agriculture 2015) at the station, \( {\overline{R}}_i \) is the expected rainfall depth from the frequency distribution method (i.e., probability distribution function, PDF) at the station, and n is the number of storm records.

The frequency analysis estimated the probable maximum rainfall depths for the six chosen return periods at the rainfall stations. These maximum depths are spatially distributed over the catchment using the IDW technique. It is a well-known interpolation technique, which resulted into spatially distributed probable maximum rainfall depths for the six different return periods (Viessman et al. 1977),

$$ P\left({x}_0,{y}_0\right)={\sum}_{i=1}^m{w}_i{P}_i $$
(2)

where \( {w}_i=\frac{\frac{1}{{{\mathbf{x}}_i}^2}}{\sum_{i=1}^m\frac{1}{{{\mathbf{x}}_i}^2}} \), xi2 = (xi − x0)2 + (yi − y0)2, P(x0, y0) is the estimated rainfall at coordinates (x0, y0), Pi is the rainfall at the given station i, wi is the station weight, and xi = (xi, yi) is the coordinates of the station.

Designing RDPs

A large dataset of detailed storm subdaily records is required for designing the RDPs. The data are grouped according to the location of the maximum intensity. Four groups of storms were formed according to the quartile number (first quartile, second quartile, third quartile, and fourth quartile). Mass curves were developed to relate the rainfall depths with their corresponding durations. The mass curves were transformed into dimensionless by dividing the individual rainfall depth of each storm by their total by depth on the y-axis and dividing the corresponding elapsed time by the total duration of the storm on the x-axis. Elfeki et al. (2014) followed this procedure and developed mass curves for Medina region. These temporal RDPs are instrumental in developing appropriate hyetographs for the location under study using the maximum rainfall depths for different return periods. Such temporal distributions are essential inputs for the hydrological modeling.

Watershed modeling

Watershed modeling was performed in WMS software in order to delineate the catchment boundary characteristics and to provide a schematic model for the HEC-HMS software and subsequent Hydraulic Engineering Center’s River Analysis System (HEC-RAS version 5.0) software. All pertinent data were obtained for different modeling procedures and applications. The DEM with a resolution of 30 m for Medina was used in WMS software for the purpose of a catchment modeling, as shown in Fig. 6. WMS generated flow directions and stream networks using the DEM, which was used in the catchment delineation. Geographic Information System (GIS) and remote sensing approaches were used for preparation of the land use data and soil type data for the TU-IU catchment. Land use and land cover maps were prepared from Image Sentinel-2 with 10 m by 10 m resolution, which was acquired from the European Space Agency (ESA) on 3 December 2016. The morphometric and hydrologic parameters were obtained using the integration of the delineated catchment, soil type classifications in the catchment, and land use data. These parameters along with the catchment delineation and schematic are needed for running the hydrological model in HEC-HMS.

Fig. 6
figure 6

TU-IU campuses and catchment superimposed over satellite image

Full size image

Hydrological modeling

All data obtained from WMS were used in HEC-HMS software for performing hydrological modeling. The rainfall-runoff modeling in HEC-HMS requires three essential inputs: (1) the meteorological data, which consisted of rainfall (obtained from the designed RDPs); (2) the basin parameters (obtained from the watershed modeling in WMS); and (3) the control inputs (user-defined inputs for the start and end time of the simulations). Based on these three inputs, the simulations in HEC-HMS were performed for different frequencies of floods using the designed RDPs. For evaluating the importance of the designed RDPs and assessing the validity of the SCS-type RDPs, it was necessary to make a comparison between the designed RDPs and the SCS-type RDPs. Therefore, the hydrologic modeling was also performed for the different frequencies of floods using the SCS type II RDPs and then simulated results were compared.

Hydraulic modeling

The application of HEC-RAS flow model requires a large number of iterative steps for numerical solutions of wave models contained within the subroutines of different modeling approaches and their different options. The different options have graphic user interface (GUI) pop-up windows used for entering different types of data that are required for hydraulic modeling, like channel geometric features, hydraulic parameters, and flow data from the hydrographs, in order to evaluate water elevations and flood inundations for different frequencies. Peak discharges for different frequencies of floods based on the designed RDPs and the SCS type II RDPs were obtained from the runoff hydrographs. These peak discharges, along with the geometric data which consisted of the TU-IU channel and their cross-sections, were used in HEC-RAS software for hydraulic modeling (Elfeki et al. 2015). The simulations were performed for different return periods based on the designed RDPs and the SCS type II RDPs. The hydraulic model in the HEC-RAS generated flood inundation depths and flood inundation widths showing the inundation area. For visualizing the flood maps, the data from the HEC-RAS were exported to GIS software. Processing of HEC-RAS data was performed in GIS software for preparing the flood inundation maps.

Lastly, statistical analyses were performed in order to compare the designed RDPs and the SCS type II RDPs using different outputs from both hydrological and hydraulic models for different flood frequencies.

Results and discussion

Frequency analysis results

Different statistical distribution techniques were applied based on the rainfall records of over 40 years. SMADA program was used for employing these techniques. RMSE values, as represented in Table 1, were analyzed for choosing the relatively best statistical distribution technique for estimating the maximum depths at station M001 and station M103. From Table 1, it is evident that two-parameter log-normal and three-parameter log-normal distribution techniques resulted in relatively small RMSE values of 7.92 and 4.56 for station M001 and station M103, respectively. Thus, these two distribution techniques were used for estimating the maximum depths for six different return periods.

Table 1 Comparison of different statistical distribution techniques based on RMSE values
Full size table

Apart from the RMSE values, probability plots of different statistical techniques (shown in Fig. 7) for the rainfall records of over 30 years validate the fact that the two-parameter log-normal and three-parameter log-normal distribution curves relatively best fitted the given records of data among the other distribution techniques.

Fig. 7
figure 7

Probability plots of different statistical distribution techniques for fitting the given daily rainfall depths at stations in the study area: M001 and M103: a, b fitting different distributions; c, d the best distribution based on RMSE

Full size image

Six return periods of 5, 10, 25, 50, 100, and 200 years are used in this study for the rigorous comparison of the designed RDPs and the SCS type II RDPs. The maximum estimated depths for six return periods based on the two-parameter log-normal and three-parameter log-normal distribution techniques at station M001 and station M103 are given in Table 2.

Table 2 Estimated maximum depths (mm) for different return periods based on the best distributions for M001 and M103 and the expected average of the maximum rainfall for the catchment
Full size table

These maximum depths of the rainfall events were distributed over the TU-IU catchment using the IDW interpolation technique. The expected maximum rainfall depths distributed over the whole catchment are presented in Table 2.

Comparison between SCS type II and the designed RDPs

The dimensionless cumulative rainfall SCS type II and designed rainfall curve for the study area, modified from Elfeki et al. (2014), are represented in Fig. 8. The cumulative rainfall depths based on the designed RDPs were obtained by multiplying maximum rainfall depths for different return periods, as shown in Fig. 9. Applications of hyetographs of the designed RDPs and SCS type II RDPs (Viessman et al. 1977) are presented in the hydrological modeling subsection under the “Results and discussion” section.

Fig. 8
figure 8

Dimensionless cumulative rainfall distribution curves for Medina region

Full size image
Fig. 9
figure 9

Cumulative rainfall depths based on the designed RDPs (a) and SCS type II (b) for TU-IU catchment

Full size image

Watershed modeling results

The catchment characteristics were obtained using the WMS software. The morphological parameters of the catchment are summarized in Table 3. GIS software and remote sensing were used to obtain the land cover characteristics and the soil type classification. Maximum-likelihood criterion was used in Environment for Visualization Images (Envi) software to obtain the land cover characteristics. This approach is a common supervised learning technique in which the known land use categories like an urban area, rocks, etc. were used as training classes and Envi software allocated cover types to each pixel in the image to which the spectral response is most similar, as shown in Fig. 6. Land cover characteristics indicated that the land consists of some vegetation, urban area, infrastructure (like roads), soil, and rocks, as shown in Fig. 10. The soil type is usually identified by the Hydrological Soil Group (HSG). The Curve Number (CN) for HSG and the corresponding land use features of the TU-IU catchment was found to be 86.1, using weighted composite CN (Hjelmfelt 1991; Ponce and Hawkins 1996), as shown in Table 4.

Table 3 Morphological parameters of TU-IU catchment
Full size table
Fig. 10
figure 10

Land use map of the TU-IU catchment

Full size image
Table 4 Land cover characteristics and the composite CN value
Full size table

Hydrologic modeling results

Hydrological modeling was performed using HEC-HMS software, which gave the peak discharges for six different return periods. The objective of the hydrological modeling was to perform rainfall-runoff modeling. The rainfall-runoff models produced runoff hydrographs which were essential inputs for the flood modeling. The input data for HEC-HMS were obtained from watershed modeling, the most important of them were the catchment schematic, drainage network, and the morphological parameters of the TU-IU catchment. Different simulations were performed in HEC-HMS, which were based on the different hyetographs for six different return periods. The RDPs for different hyetographs for six return periods were grouped into two types. One of them was based on the designed RDPs and the other was based on the SCS type II RDPs. The simulated results from the designed RDPs and the SCS type II RDPs are presented in Fig. 11.

Fig. 11
figure 11

Runoff hydrographs based on the designed RDPs (a) and SCS type II RDPs (b) for different return periods

Full size image

The peak discharges from the hydrological modeling for both types of the RDPs are plotted in Fig. 12. Peak discharges, volumes, and time at peak discharges are summarized in Table 5, which were used as inputs to the hydraulic modeling of the TU-IU catchment. By comparing the designed RDPs and the SCS type II RDPs, it is evident that the peak discharges are overestimated for the SCS type II RDPs. The difference in peak discharges increases with respect to the recurrence interval. This leads to the inaccurate and expensive design of the infrastructure in the TU-IU catchment. Another important difference is the starting time of the runoff hydrographs. The runoff hydrographs are delayed for the SCS type II RDPs as compared to the designed RDPs. The approximate starting time for the hydrographs based on the SCS type II RDPs is after 1 h from the rainfall event, while the designed RDPs indicate that runoff hydrographs are immediately started as soon as the storm events occur. The time at peak discharges in Table 5 indicates that the peak discharge for the SCS type II RDPs relatively occurred at around 1 h later as compared to the designed RDPs. The negative volume difference means that the volumes for the SCS type II RDPs are lesser than that of the designed RDPs (the designed RDPs values are always subtracted from the SCS type II RDPs values). The runoff hydrographs based on the designed RDPs show a more realistic representation of the excess rainfall which are proven from the flash flood events that happened in the arid regions immediately after the torrential storm event (Elfeki et al. 2014). Thus, the hydrological modeling establishes the fact that the use of SCS type II RDPs in the arid regions, like Medina, is unjustified, while the designed RDPs show the realistic features of the actual events in arid regions.

Fig. 12
figure 12

Comparison of peak discharges for two different types of RDPs

Full size image
Table 5 Comparisons of time at peak discharges, volumes (runoff hydrographs), and peak discharges between the designed RDPs and the SCS type II RDPs
Full size table

Hydraulic modeling results

Hydraulic modeling was performed using HEC-RAS software. The peak discharges from runoff hydrographs were used as the input data. Numerous simulations were performed based on the peak discharges for two different RDPs. From hydraulic modeling application, flood inundation widths, flood inundation maximum depths, minimum depths, average depths, and flow velocities were obtained for the two different types of RDPs.

Longitudinal profiles of flood flow for both types of RDPs from the hydraulic modeling are compared in Fig. 13. For a more detailed analysis, the different parameters like flood inundation depths, widths, areas, and flow velocities are compared in Fig. 14 and for their quantitative analysis. Figure 14 indicates increasing trend for the flood inundation depths, widths, areas, and velocities, thus representing an overestimation of these variables when SCS type II RDPs were used. From Fig. 14, it is evident that the differences of inundation widths, areas, and velocities are more significant as compared to the inundation depths. However, the pattern of velocity profiles along the channel do not show variations, as shown in Fig. 15. The velocities in the channel fluctuate as expected where the fluctuations approximately range from 1.9 to 10 m/s, for both types of RDPs (velocities of the SCS type II RDPs are comparatively higher). The flood inundation mappings of the TU-IU channel are compared in Figs. 16 and 17 for the two different types of RDPs. The mappings are plotted to cover the whole catchment; therefore, the differences may not be significantly evident from the figures, and in addition to this, the inundation patterns are the same for all mappings. The differences in the inundation depths are illustrated in their respective legends.

Fig. 13
figure 13

Comparison of the longitudinal flow profiles along the channel of the designed RDPs (a–c) and the SCS type II RDPs (d–f) for the return periods of 5, 50, and 200 years

Full size image
Fig. 14
figure 14

Comparisons of a flood inundation depths, b widths, c flood areas, and d velocities between the two different types of RDPs

Full size image
Fig. 15
figure 15

Comparison of velocity profiles along the channels between a the designed RDPs and b the SCS type II RDPs for different return periods (Vel Chnl—flow velocity in the channel)

Full size image
Fig. 16
figure 16

Flood inundation mapping of the Taibah University channel, where (a), (b), and (c) represent the return periods of 5, 50, and 200 years for the designed RDPs, and (d), (e), and (f) represent the return periods of 5, 50, and 200 years for the SCS type II RDPS, respectively

Full size image
Fig. 17
figure 17

Flood inundation mapping of the Islamic University channel, where (a), (b), and (c) represent the return periods of 5, 50, and 200 years for the designed RDPs, and (d), (e), and (f) represent the return periods of 5, 50, and 200 years for the SCS type II RDPS, respectively

Full size image

Summary and conclusions

Common practices in arid regions indicate that the SCS type II RDPs are used by the hydrologists. However, there is no scientific justification available for such practices. In this study, the two different types of RDPs were analyzed for the TU-IU catchment in Medina, Saudi Arabia, which is an arid region. One of them is the designed RDPs and the other is the SCS type II RDPs. First, a rigorous methodology was used for accurately comparing the two patterns. The methodology involved collection, organization, and evaluation of the rainfall records at the two stations (station M001 and station M103). Different types of frequency analyses were performed for statistically distributing the maximum rainfall depths for six chosen return periods (5, 10, 25, 50, 100, and 200 years). SMADA program evaluated the different distribution techniques using the RMSE criterion. It is indicated that the two-parameter and three-parameter log-normal techniques generate the smallest RMSE values as compared to the other distribution techniques at station M001 and station M103, respectively. These two techniques were used for estimating the probable maximum rainfall depths which were later distributed over the whole TU-IU catchment using the IDW interpolation technique. The maximum values of rainfall depths were distributed over the storm duration using the designed RDPs from the dimensionless RDPs presented by Elfeki et al. (2014). These RDPs for different return periods were used for the hydrological modeling in HEC-HMS which were then compared with the SCS type II RDPs. For hydrological modeling, the watershed modeling was performed in WMS, in which the morphological characteristics were obtained together with the catchment schematic. GIS and remote sensing were used for obtaining the CN by integrating the land use and soil type characteristics of the catchment. For peak discharges, the SCS type II gave about 18% higher values than the designed RDPs for all return periods. The time to peak for the SCS type II is longer by about 1 hour, on the average, when compared with the designed RDPs. This conclusion is vital for the mitigation measures. The results of flood inundation depths, widths, areas, and velocities were compared for both types of RDPs, which indicated that the SCS type II RDPs overestimated these variables by about 3.1, 7.7, 13.3, and 5%, respectively. Hence, using the SCS type II RDPs for designing the hydraulic structures will be inaccurate, unjustified, and uneconomical. Therefore, the designed RDPs are the most suitable for the arid regions and should be employed in designing new water resources projects in the arid regions.