3.1 Rainfall Frequency Analysis
In this study, the rainfall frequency analysis was determined by four commonly used probability distributions of hydrology. The chosen methods were Normal, Log-Normal, Extreme Value Type-I (Gumbel), and Log Pearson Type III. However, the ERA manual recommended Gumbel and log person type III for rainfall frequency analysis. The summary of Extreme rainfall for all methods for different return periods is as tabulated in Table 1.
To identify which distribution fits the theoretical probability distribution, the GOF test was conducted using Easy Fit 5.6 professional software. As a result, the Gumbel distribution fits better than the rest of the method. In this study, in addition to the Easy-fit software, the correlation coefficient was used to identify the best fit distribution. So, Gumbel has a better R2 than the other methods which is 0.998.
3.2 Intensity-duration-frequency curve (IDF)
The IDF curve was developed from a 24-hour rainfall data of 32 years duration that is from 1987 to the 2018 year, which was obtained from the Ethiopian Meteorological Agency were the gauge station located at Ambo Agricultural center in Ambo town. Since the shorter rainfall data is not available, the Reduction equation was used to calculate the shorter rainfall data needed for the development of the IDF Curve. The rainfall intensity was calculated for 5minutes intervals up to 180 minutes, and its result was shown in the tables 2. From the rainfall intensity result table 2, the Intensity duration frequency curve was developed for the Ambo station as shown in figure 2.
3.3 Peak discharge determination using the rational method
Since all sub-catchment areas of this study are less than 50ha it’s recommended to use rational methods to determine the peak runoff from the sub-catchment. The weighted runoff coefficients were determined from the land-use composition for each sub-catchment using the runoff coefficient for each land use as discussed in the methodology part. Time of concentration consists of inlet time plus time of flow in a channel as discussed under the methodology part of this study and its result as shown in a table for each sub-catchment. So, using the above input parameters, the peak discharge from the catchment was determined using the rational method as shown in table 3.
3.4 Stormwater management model simulation result
The performance of the stormwater drainage network of the study area was assessed using the stormwater management model. To assess the performance of the stormwater drainage system of the town, the software needs input data for modeling runoff quantity and others. The details of input data used for physical catchment were collected using field surveying and other adopted from different hydrologic kinds of literature. As well as the detail of conduits properties and Junction properties were measured at the field. The maximum, minimum infiltration rate and decay constant values were adopted from (Gilliom et al. 2019; Liu et al. 2022) depending on the study area's soil type.
Finally, after the input parameters of the model were collected, processed, and inserted, the model simulates successfully. Then, once the SWMM is successfully simulated, the simulated result were discussed in the following.
3.4.1 Drainage network mapping
The total study area is divided into 40 sub-catchments and the network consists of 73 nodes or junctions, 8 outfalls, 75 conduits links, and one rain gage station. For this study the sub-catchments area is denoted as Sc, the nodes as J, the conduit as C, and the rain gage denoted as RG. The sub-catchment was classified depending on the area. As seen from the modeling map, most of the sub-catchment were less than 25ha and only two sub-catchments were greater than 25ha. The node in the modeling map was classified depending on the inverted elevation. Since invert elevation for the whole junction is greater than 100, it is displayed in red color. Again the link was classified depending on the link depth. From the modeling map, the majority of link depth ranges from 0.5 to 1m, whereas six links were less than 0.5m. The total model map of the study area includes sub-catchments, nodes, and the link was shown in figure 3.
3.5 Calibration and validation of the model
Modeling study typically requires calibration and validation to evaluate model performance. For SWMM, some model parameters are directly measured and their value was fixed through onsite investigation. Others which are the only conceptual representation of catchment features necessarily be determined through a trial and error process until a satisfactory matching is obtained between the model output and measured data(Perin et al. 2020). Since there is no measured data in the study area, the calculated runoff using the rational methods was used as measured data for fixing the sensitive parameters of the model, and Sensitive parameters value range for SWMM hydrology and hydraulic parameters was given in table 4 according(Akdoğan and Güven 2016).
3.5.1 Model performance evaluation
The performance of SWMM was evaluated using runoff calculated using the rational method and the simulated result of the software. According to the calculated and simulated value of discharge for a 10year return period were RNS value is 0.997 and R2 is 0.997. Since the RNS and R2 are 0.99 it shows that the model gives a good result.
3.5.2 Comparison of discharge results
A comparison was done between the discharge determined by rational methods for each sub-catchment and the discharge from the SWMM or the simulated results.
From figure 4; the correlation between both results is 0.998 which implies that the simulated result with SWMM well matches the rational method. So, the calibration and validation of hydrological parameters gave an excellent result.
3.7 Water depth and flow in the canal
The water depth and flow in a canal were also simulated to assess the adequacy of the canal. The water depth simulation results for a canal were presented as follows.
3.7.1 Water elevation profile toward outfall 1
Figure 5; shows that the water profile in canals 3 and 4 is the design water depth. That means canals 3 and 4 are sufficient to carry the generated runoff. So there are no flooding injunctions 2, 3, and 6.
3.7.2 Water elevation profile toward outfall 2
From the simulation result, figure 6; shows that the water profile at junction 5 is insufficient to carry the generated runoff from the sub-catchment and there is a flood at junction 5 and outfall 2.
3.7.3 Water elevation profile toward outfall 3
As the simulation result indicates, the drainage canal toward outfall 3 means junction 9 to outlet 3 is sufficient to carry the generated runoff. So, there is no flooding in canal 9 as shown in figure 7.
3.7.4 Water elevation profile toward outfall 4
As shown in figure 8 the drainage canal toward outfall 4 means the junction 10, 11, 19 to outlet 4 is sufficient to carry the generated runoff. So, there is no flooding in canals 10, 11, 19, and 23.
3.7.5 Water elevation profile toward outfall 5
From the simulation result of SWMM, the majority of the canal that dispose of stormwater toward the outfall 5 is surcharged. As shown in figure 9, the junctions 14, 15, 16, 17, 24, 30, 32, and outfall 5 are highly flooded, whereas the rest of the junctions are sufficient to carry the generated runoff from the sub-catchment.
3.7.6 Water elevation profile toward outfall 6
As the simulation result indicates, the flow level in drainage canal 45 is highly flooded. As shown in figure 10, both junctions 38 and 43 are insufficient. Again from the simulation, Junction 45, 46, 50, and 51 are sufficient to safely dispose of the generated runoff from the sub-catchment
3.7.7 Water elevation profile toward outfall 7
From the simulation result as shown in figure 11, the drainage canal toward outfall 7 means that junction 57, 60, and 69 is insufficient to carry the generated runoff. So, there is flooding in canals 58 and 70. Except for the junction listed above, it's sufficient or there is no surcharge in the canal.
3.7.8 Water elevation profile toward outfall 8
Figure 12, shows the water flow depth in the canal toward the outfall 8. As can be from the figure below that, junctions 73 and 72 which connect to canal 73 are highly flooding due to insufficient canal depth. But, both canals 74 and 75 are sufficient to carry the generated runoff. So there are no flooding injunctions 71 and 79.