This study assessed the factors affecting the growth and survival of microorganisms in a small horizontal subsurface flow constructed wetland (HSSF CW) treating stormwater runoff from highly impervious road and parking lot through long-term monitoring from 2010 until present. The HSSF CW facility consisted of sedimentation or pre-treatment zone, vegetation zone, and effluent zone, and employed filter media including bio-ceramics, sand, gravel, and wood chips. Results showed that flow reduction in the wetland through filtration and sedimentation played an important part in the overall performance of the HSSF CW. In addition, vegetation growth was found to be affected by pollutant and stormwater inflow in the HSSF CW. Vegetation near the outflow port exhibited greater growth rates by about 6.5% to 64.2% compared to the vegetation near the inflow port due to the less stormwater pollutant concentrations via filtration mechanism in the plant or media zone of the HSSF CW. The pollutant inflow from road and parking lot played an important role in providing good environment for microbial growth especially for the dominant microbial phyla including Proteobacteria, Actinobacteria and Acidobacteria in the HSSF CW. The findings of this research are useful in understanding treatment mechanisms and identifying appropriate design considerations for HSSF CW.

Twenty-three rainfall events were monitored to determine the characteristics of the stormwater runoff entering a rain garden facility and evaluate its performance in terms of pollutant removal and volume reduction. Data gathered during the five-year monitoring period were utilized to develop a deep learning-based model that can predict the concentrations of Total Suspended Solids (TSS), Chemical Oxygen Demand (COD), Total Nitrogen (TN), and Total Phosphorus (TP). Findings revealed that the rain garden was capable of effectively reducing solids, organics, nutrients, and heavy metals from stormwater runoff during the five-year period when hydrologic and climate conditions have changed. Volume reduction was also high but can decrease over time due to the accumulation of solids in the facility which reduced the infiltration capacity and increased ponding and overflows especially during heavy rainfalls. A preliminary development of a water quality prediction model based on long short-term memory (LSTM) architecture was also developed to be able to potentially reduce the labor and costs associated with on-site monitoring in the future. The LSTM model predicted pollutant concentrations that are close to the actual values with a mean square error of 0.36 during calibration and a less than 10% difference from the measured values during validation. The study showed the potential of using deep learning architecture for the prediction of stormwater quality parameters entering rain gardens. While this study is still in the preliminary stage, it can potentially be improved for use in performance monitoring, decision-making regarding maintenance, and design of similar technologies in the future.

The efficiency of nature-based facilities is mostly evaluated in terms of their pollutant removal capabilities; however, apart from water purification functions, constructed wetlands also perform ecological, cultural, and environmental education functions. A multi-criteria performance index was developed in this study to evaluate the overall efficiency of constructed wetlands. A total of 54 constructed wetlands installed across South Korea were monitored to evaluate the pollutant removal performance of the facilities. A comparison between the conventional pollutant removal-based evaluation and the developed multi-criteria index was also performed to determine the key changes in the results of evaluation when different methods are employed. Among the different types of wetlands studied, hybrid systems were found to be the most effective in terms of pollutant removal due to their complex configurations and functions. Newly constructed treatment wetlands have unstable performance and thus, a stabilization period ranging from two to five years is required to assess the facility’s pollutant removal capabilities. As compared to the conventional pollutant removal-based efficiency evaluation, the comprehensive evaluation method provided a more strategic tool for identifying the facilities’ strengths and weaknesses. Generally, the multi-criteria performance index developed in this inquiry can be utilized as a general tool for evaluating the sustainability of similar nature-based facilities.

Evaluating the functionality of small and decentralized low-impact development (LID) technologies often requires extensive labor, time, and costs for water quality analysis. In order to reduce these in an infiltration trench in South Korea, monitoring data gathered over a period of 8 years were used to determine its long-term performance, establish a stormwater quality estimation model, and develop a comprehensive evaluation tool. Our findings show that the infiltration trench can treat up to 90% of the stormwater runoff from a paved road but would require annual maintenance to minimize the reduction in infiltration capacity. The facility was able to remove an average of 83% of total suspended solids (TSS), 75% of biochemical oxygen demand (BOD), 80% of chemical oxygen demand (COD), 76% of total nitrogen (TN), and 79% of total phosphorus (TP), with the highest removal efficiencies observed after maintenance was conducted. Rainfall depth and air quality parameters (i.e., PM_2.5 and PM₁₀) were found to be positively correlated with TSS, COD, TN, and TP. These parameters were then used to develop a model for the estimation of influent stormwater quality, which can help in estimating the effluent water quality based on the average removal efficiencies. Furthermore, a comprehensive evaluation tool considering indicators such as treatment efficiency, cultural benefits, and facility and operating conditions was established through the analytic hierarchy process (AHP). Aside from determining the facility’s overall efficiency, this can also serve as a diagnostic tool to identify whether maintenance is needed or not. While atmospheric and hydrological characteristics differ in different regions, and the results may vary if applied in other facilities, this study can serve as a guide to the effective and efficient evaluation of similar stormwater management facilities in South Korea.

Low impact development (LID) technologies mimic the natural water cycle through the physico-chemical and biological interactions of plants, filter media and soil, and microorganisms, thereby reducing the release of pollutants. In LID facilities, plants carry out photosynthesis, facilitate microbial growth, and uptake pollutants contained in stormwater runoff. However, de-icers (CaCl₂) used to melt snow during winter slow the growth of plants and even increase plant mortality. In addition, de-icers change the soil structure, causing changes in soil content and affecting the growth of plants and microorganisms. Therefore, this study examined the effects of CaCl₂ on the resistance of plants, the removal efficiency of non-point source pollutants, and water circulation. The mortality rate of the tree and shrubs caused by CaCl₂ was found to be in the order of Rhododendron indicum > Spiraea prunifolia var. simpliciflora > Metasequoia glyptostroboides. For herbaceous plants, mortality rate was in the order of Pratia pedunculata > Aquilegia japonica > Tagetes erecta > Sedum makinoi aurea > Hosta longipes > Dianthus chinensis > Acorus gramineus > Liriope platyphylla. In addition, it was found that the amount of chlorophyll decreases with high concentrations of CaCl₂. The findings of this research will be useful for plant selection considering CaCl₂ concentrations applied to paved areas during the winter.

Most of the studies about stormwater low-impact development technologies (LID) used generalized observations without fully understanding the mechanisms affecting the whole performance of the systems from the catchment to the facility itself. At present, these LID technologies have been treated as black box due to fluctuating flow and environmental conditions affecting its operation and treatment performance. As such, the implications of microbial community to the overall performance of the tree-box filter (TBF) were investigated in this study. Based on the results, summer season was found to be the most suitable season for microorganism growth as greater microorganism count was found in TBF during this season compared to other seasons. Least microorganism count was found in spring which might have been affected by the plant growth during this season since plant penology influences the seasonal dynamics of soil microorganisms. Litterfall during fall season might have affected the microorganism count during winter as, during this season, the compositional variety of soil organic matter changes affecting growth of soil microbial communities. Microbial analyses of soil samples collected in TBF revealed that the most dominant microorganism phylum is Proteobacteria in all the seasons in both inlet and outlet comprising 37% to 47% of the total microorganism count. Proteobacteria is of great importance to carbon, sulfur, and nitrogen cycling in soil. Proteobacteria was followed by Acidobacteria, Actinobacteria, and Chloroflexi which comprises 6% to 20%, 9% to 20%, and 2% to 27%, respectively, of the total microorganism count for each season. Each microorganism phylum was found to have varying correlation to different soil chemical parameters implying the effects of these parameters to microorganism survival in LID technologies. Depending on the target biogeochemical cycle, maintaining a good environment for a specific microbial phylum may be decided. These findings were useful in optimizing the design and performance of tree box filters considering physical, chemical, and biological pollutant removal mechanisms.