The characteristics of rainfall runoff pollution and its driving factors in Northwest semiarid region of China - A case study of Xi'an

Highlights

• Spatio-temporal trend and driving factors of rainfall runoff pollution were detected.

• Rainwater quality was relatively good but road runoff quality was poor.

• TN decreased in rainwater but increased in roof and road runoff.

• Technology innovation was the major contributor to the improvement of water quality.

• Population scale effect sharply increased due to the implementation of new policy.

Abstract

With the effective control of point source pollution, rainfall runoff pollution has become the main source of water pollution in Xi'an. Understanding the characteristics and driving factors of rainfall runoff pollution would provide theoretical foundation for urban rainfall runoff pollution control. In this study, a total of 32 rainwater samples, 604 roof runoff samples and 608 road runoff samples obtained from literature and a total of 35 rainwater samples, 127 roof runoff samples and 70 road runoff samples collected by our group were mixed to analyze the rainfall runoff pollution from 2008 to 2019. The Mann-Kendall and Logarithmic Mean Divisia Index (LMDI) method were used to determine the trend of pollution and the driving factors of water quality changes. The results showed that the major pollutant from rainwater was nitrogen, while the main pollutants from roof and road runoff were COD and SS. The rainwater quality during the study period was “clean to slightly polluted”. The roof runoff quality of most years (67%) was “clean to slightly polluted”. The road runoff quality was poor: 22% was “moderately polluted”, and 45% was “heavily polluted”. The concentration of pollutants except COD in rainwater showed a decreasing trend, while the trend of pollutants in roof and road runoff was not completely consistent with that in rainwater. NH₃-N showed strong positive correlation with TN in roof runoff, which indicated common sources of these pollutants. There was a significant correlation between SS and COD in road runoff, and between SS and TP in roof runoff, suggesting SS was an important carrier of COD and TP. Technology innovation was the dominant factor affecting water quality, followed by industrial structure. Economic development and population scale contribute negatively to water quality improvement, and there was a sharp increase in the population scale effect in 2017.

Introduction

In recent years, the total area of impermeable surface has increased dramatically due to rapid urbanization, leading to high risks of flooding and rainfall runoff pollution (Brabec et al., 2002; Todeschini, 2016). At the same time, rainfall runoff, which contains numerous pollutants (such as nitrogen, phosphorus, heavy metals, and organic substances), entered receiving water bodies and lead to water quality deterioration (Liu et al., 2018; L.X. Zhang et al., 2018; P.P. Zhang et al., 2018; Qin et al., 2016; Wang et al., 2016). Since roof runoff and road runoff contributed 5.35% and 69.24% particles to urban receiving water, respectively (Ma et al., 2018), rainfall runoff has become a primary contributor to receiving water pollution (Wijesiri et al., 2015). To reduce the urban non-point source pollution for source control and achieve high-quality sustainable development of cities, the rainfall runoff pollution characteristics and its influencing factors need to be identified.

The variation of rainfall runoff pollution has been addressed in many previous studies. The pollutants concentrations in runoff were high in the northwest semiarid region of China, and most pollutants concentrations exceed Class V of Surface Water Environmental Quality Standards (GB3838-2002) (Y. Chen et al., 2017; Wang et al., 2019; Wang et al., 2017; Wu et al., 2018; Xu et al., 2018). Furthermore, some studies showed that there were differences in runoff pollution of different functional areas, and the average EMC values from high to low were business district, traffic area, residential area, institutes and colleges district (Gong et al., 2019). However, these studies are limited to a small area and a short time series, which ranging from a few months to a few years. It is challenging to capture the dynamic nature of urban environments and long-term rainfall runoff pollution trends.

Rainfall runoff pollution is associated with various factors including climate (Zuo et al., 2016; dos Santos et al., 2019), rainfall characteristics (Lee et al., 2011; Yuan et al., 2018), air quality (Han et al., 2019), traffic (Brigitte et al., 2010) and land use (Zhang et al., 2010; Wang et al., 2013), etc. The amount of pollutants build-up on road surfaces could plateau when antecedent dry period (ADP) was greater than seven days (Mahbub et al., 2011), and pollutant concentrations in road runoff are significantly influenced by rainfall intensity (Ma et al., 2018). Furthermore, Men et al. (2018) found that source contribution of heavy metals in road runoff changes with the variation of human activities. However, research into how rainwater and rainfall runoff quality responds to human socio-economic activities (e.g. population growth, economic development, and energy consumption) is currently limited. To date, there are numerous studies on socio-economic driving factors of environmental pollution by applying the Logarithmic Mean Divisia Index (LMDI) method, but most of them are focused on air pollution. For example, Dong et al. (2019) found that the economy is the main driving force for global carbon emissions. Wang et al. (2020) reported the major driving forces of the nitrogen oxides intensity are the clean energy penetration and thermal power generation efficiency. In this study, the LMDI decomposition model was applied to identify the key drivers of rainwater and rainfall runoff quality change, which is helpful to develop strategies on pollution control.

Therefore, this study aims to understand the characteristics and driving factors of rainfall runoff pollution based on the long time sequence rainfall runoff quality data (from 2008 to 2019) in Xi'an. The COD, NH₃-N, TP, TN, and SS with sufficient statistical samples were chosen to: (1) understand the characteristics and the pollution level of rainwater and rainfall runoff, (2) describe long-term water quality variation trends of rainwater and rainfall runoff by using the Mann-Kendall method, and (3) identify the correlation of each pollutants in rainwater and rainfall runoff, and their main driving factors. The understanding of rainfall runoff pollution based on Xi'an can provide theoretical foundation for other cities in semiarid regions.