Evaluating the effects of climate change on urban runoff based on CMIP6 models (Case study: district 10 of Tehran municipality)

Introduction
In recent years, the changes in the intensity and frequency of precipitation and the occurrence of severe floods and droughts have prompted decision-makers to consider the effects of climate change in their plans. Due to the existence of impervious areas in urban environments, a more significant part of precipitation is convertedto runoff, and the changes in precipitation patterns resulting from climate change can affect the performance of drainage systems. On the other hand, with the change in precipitation pattern, the amount of pollutants washed from the surface is changed, and in this way, the quality of runoff is also affected. Nowadays, coupled atmosphere-ocean general circulation models (AOGCMs) are considered the most advanced and reliable tools for simulating climate change. Recently, a coupled model intercomparison project Phase 6 (CMIP6) hasbeen introduced as the latest version of AOGCMs, which can simulate future periods with high accuracy. The sixth assessment report evaluates the changes in climate variables by combining Representative Concentration Pathway (RCP) and Shared Socioeconomic Pathway (SSP) scenarios. According to this report, in addition to covering different climates, future scenarios should also consider the socio-economic aspects of development. The CMIP6 models have higher spatial resolution than the models of previous reports. A review of the research background in the assessment of climate change effects on precipitation and runoff shows that most of the studies have been conducted using the models and scenarios of the fifth and earlier reports, and only a few of them have included the scenarios of the sixth report in their evaluations. In this regard, our research evaluated climate change effects on urban runoff based on the CMIP6 models’ predictions.
 
Materials and Methods
District 10 of Tehran municipality is selected as the case study. This region is located in the south of Tehran and has an area of about 800 ha. Due to its high population density, lack of enough green space, and a high percentage of impervious areas, runoff management is a priority for this region. Moreover, the presence of agricultural land highlights the need for runoff quality management in the study area. The MehrAbad synoptic station is the nearest to the case study, and its observation data during the base period is gathered for evaluating the changes in hydrological variables under climate change. For implementing the methodology, different CMIP6 models were first assessed, and those with high performance in precipitation prediction in the historical period were selected. Their projections under SSP1-2.6 and SSP5-8.5 for the future (2021-2050) were downscaled using the LARS-WG model. Then, the statistical method was employed for disaggregating the LARS-WG’s daily output into 6-hour design precipitation. In the following, maximum and minimum values of precipitation were determined as optimistic and pessimistic scenarios, respectively, and the stormwater management model (SWMM) was implemented for simulating the runoff under these scenarios. The SWMM subdivides the watershed into sub-watersheds and utilizes three primary processes for runoff quality and quantity simulation. First, generated runoff is calculated by determining the hydrological characteristics of the sub-watersheds in the hydrologic process. Then, the canals route runoff to the outlet using a hydraulic process. In the quality process, the runoff quality is simulated using build-up and wash-off equations. In our research, the study area was subdivided into 84 sub-watersheds. Changes in runoff volume, peak flow, and total suspended solid (TSS) concentration at the watershed outlet were evaluated under climate change. Correspondingly, for assessing the performance of the drainage system, the changes in the flooded volume of the system were quantified.
 
Results and Discussion
According to the results, in all models under SSP5-8.5, monthly precipitation will increase in January, February, and March and decrease in August and September. Also, under SSP1-2.6, the precipitation trend is predicted to fall in September. The highest increase in precipitation compared to the base period is related to August under SSP1-2.6. In addition, with a decrease of 37.4 %, the highest reduction in precipitation is associated with February. The most evaluated prediction uncertainty is related to August under SSP1-2.6. This month, precipitation changes range from +226.31 to -18.34 % compared to the base period. Also, predictions on an annual scale do not show a specific trend. The changes in annual precipitation vary from a -9.8 % decrease to a 5.4 % increase compared to the base period. Then, by analyzing the 6-hour rain, the predicted values of HADGEM3-GC31-LL and CMCC-ESM2 were identified as the highest and lowest values, respectively. The 6 h rain with 5 and 10-year return periods under the pessimist scenario will increase by 31.4 and 26.8 % and decrease by 2.5 and 11.3 % under the optimistic scenario, respectively. The results of performing SWMM under a pessimistic scenario showed that in the return periods of 5 and 10 years, runoff volume would increase by 25.2 and 20.7 %, and TSS concentration will decrease by 21.4 and 18.2 %, respectively. Besides, in this scenario, the flooded volume of the basin increases to 42.12 %. Performing SWMM under an optimistic scenario revealed that with the reduction of precipitation compared to the base period, in the return period of 5 and 10-years, the runoff volume will decrease by 2.2 and 8.3 %, and the TSS concentration will increase by 2.5 and 10 %, respectively.
Conclusion
Performing SWMM under an optimistic scenario shows that with the decrease of 6-hour design precipitation, the quantitative parameters (runoff volume and peak flow) decrease, and TSS concentration increases at the watershed outlet. Furthermore, under the pessimistic scenario, quantitative parameters increase, and TSS concentration decreases with the increase in precipitation. More examination revealed that despite the decline in precipitation, the number of flooded nodes remained constant under optimistic scenarios indicating the drainage system’s vulnerability even under base-case rain and a little less. Moreover, the increase in flooded volume and the number of flooded nodes under the pessimistic scenario make it necessary to utilize management strategies to improve the runoff collection systems’ performance under climate change. In this regard, low-impact development (LID) practices can be used as a climate change adaptive approach in future works.