Attribution of climate change and human activities to urban water level alterations and factors importance analysis in Central Taihu Basin
Introduction
Climate change and human activities are considered to be two major factors behind altering hydrological processes (Dey and Mishra, 2017). Warming temperatures and changes in the amount, intensity, and form of precipitation have a global impact on environment with significant implications on water resources (Arnell, 1999; Kundzewicz et al., 2008; Tramblay et al., 2012). Human activities, such as land use/cover change, hydraulic engineering construction and urbanization, exert multiple pressures on the hydrologic cycle (Shuster et al., 2005). In addition, human activities also interact with climate change (Olson et al., 2008; Berckmans et al., 2019). The superimposed effects of climate change and human activities are complex, especially in urbanization regions. Therefore, quantifying the effects of climate change and human activities on hydrological processes is valuable for strategic water resources management. In this study, climate change refers to the variations of climate factors, such as precipitation, temperature, etc., which can be due to both natural variability and anthropogenic influences. Human activities refer to the human actions that causing regional to local-scale change to the hydrological cycle.
Taihu Basin is one of the typical regions with the fastest urbanization speed and the highest urbanization level in China, which belongs to Jiangsu Province, Zhejiang Province and Shanghai City in administration. The approximately 80% of the total area in the basin is characterized by broad flat geomorphology and lowland plain river networks (Liu et al., 2013). Moreover, due to the abundant marine monsoon precipitation, Taihu Basin is susceptible to threats from flood risk (Zhang et al., 2008). Thus, attribution of climate change and human activities to water level alterations in Taihu plain could provide supportive information for flood operation.
In general, the methods of assessing the impacts of climate change and human activities on hydrological processes mainly include scenarios simulation (Li et al., 2009; Franczyk and Chang, 2009), elasticity analysis (Song et al., 2019) and residual analysis (Wang et al., 2012; Wang, 2014). Scenarios simulation refers to change land use patterns or meteorological data at a time and compare the hydrological model outputs of different scenarios to quantify the influences of climate change and human activities. The advantage of this method is that it is based on stronger physical mechanism and can reflect the inhomogeneity of climate and land use. Nevertheless, it is difficult to apply to larger and complicated basins. Elasticity can be defined as the ratio of proportional change in analysis object like streamflow and water level, and proportional change in relative factors like precipitation, temperature and impervious rate (Ma et al., 2010; Yang and Yang, 2011). This approach can access the impact of representative single factor on analysis object change. However, it is more suitable for estimating the contribution of climate factors, which is limited in the selection of human activities factors. Residual analysis is a common type of quantitative analysis that widely used to quantify the contribution of climate and human activities on the Normalized Difference Vegetation Index (NDVI) (Evans and Geerken, 2004; Zheng et al., 2019) and runoff (Fan et al., 2010). This method is usually based on the assumption that the analysis object is closely correlated with climate factors in one period, i.e., baseline period (Wang et al., 2015). In other words, human activities have no significant impact on the analysis object of analysis in this stage. Analysis object are regressed against climate factors and the resulting model is used to predict the object values for the other stages. Then the residual is the difference between the observed object and predicted object, which indicates that the changes in object are caused by human activities. This approach is simple and can effectively separate the impacts of climate change and human activities.
A large number of hydraulic structures, such as pump, sluices and dikes, have been constructed in Taihu Basin since the 1980s and the total number of hydraulic structures has reached 2468 in the 2000s (Wang et al., 2019a). Due to the complex river network and large amount of hydraulic structures, it is difficult to establish an accurate hydrological model. Consequently, residual analysis was used to quantify the impacts of climate change and human activities on water level in this study. Obviously, the regression model is a key part in residual analysis. However, the previous studies have mostly performed linear model (Li et al., 2017; Liu et al., 2018), which may not explain the relationship between analysis object and multiple factors well. The use of machine learning algorithms can bring significant advantages to both understanding and predicting the climate (Koc and Acar, 2021). Shi et al. (2020) evaluated the contributions of climate factors and human activities to vegetation greenness changes by RF regression with residual analysis. Wang et al. (2019b) distinguished the impacts of climate change and human activities on water level using the multilayer perceptron artificial neural network and residual analysis. In view of the above problems, more nonlinear regression models needed to be applied to obtain more efficient and accurate method.
The main goal of this study is to quantify the impacts of climate change and human activities on water level of Central Taihu Basin. The other objective of this study is to analysis the importance of climate factors (mainly precipitation, temperature and relative humidity) and human activities factors (daily water abstraction and drainage and daily maximum discharge of main sluices), which will be helpful for improving our understanding of important factors and providing reference for dynamic monitoring of flood control.
The water level series of study area from 1961 to 2018 was split into two subseries, i.e., baseline period and altered period, from a break year determined comprehensively by Pettitt test, rank-sum test and Mann-Whitney-Pettitt test. Linear regression and two machine learning models were applied to analyze the relationship between water level and multiple factors under baseline period, and the best one was selected to predict water level during altered period. Then the contributions of two driving factors on water level at monthly, seasonal and annual scale were separated by residual analysis. Finally, the importance of all climate and human activities factors were analyzed by RF based on daily data series from 2008 to 2018.
Section snippets
Study area
Taihu Basin is located in the middle region of the Yangtze River Delta (Fig. 1). Taihu Basin accounts for 0.4% of the total land area of China, while its GDP accounted for about 9.8% of the whole country in 2019, which is one of the highly urbanized areas of China (Taihu Basin Authority, 2020). The Central Tainhu Basin, encompassing the cities of Suzhou, Wuxi, and Changzhou, is the most typical region in the river network plain of Taihu Basin with numerous criss-crossing waterways and scattered
Water level variations
The annual water level series showed an abrupt change at around 1990, 1994 and 1990 using the Pettitt test, rank-sum test and Mann-Whitney-Pettitt test, respectively. Based on the most consistent principle, the most probable change point was determined to be 1990, a time that corresponding to the start point of rapid and massive construction of hydraulic structures in Central Tainhu Basin (Wang et al., 2019b). Then the study period was split into two parts whereby 1961–1989 was regard as the
Discussion
The importance of thirt-eight factors to daily mean water level was analyzed by RF in this study and the top five important factors (DTL, DDmax of WY, DDmax of ZJG, CAP of five days and DDmax of SYW) were recognized. The time series of the five factors and daily mean water level were shown in Fig. 8. Obviously, the variation of daily mean water level was basically consistent with the fluctuation of DTL, indicating that the DTL was the most relevant factor to daily water level among all factors.
Conclusion
This study used residual analysis to quantify contribution of climate change and human activities on the nonstationary water level alterations at monthly, seasonal and annual scale from 1990 to 2018 in Central Tainhu Basin. The results indicated that human activities were the dominant influence on the water level alterations at different time scales during altered period in the study area. To further investigate the impacts of climate change and human activities on water level alterations at a
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
This study was supported by the National Key Research and Development Program of China (Grant No. 2018YFC1508200 and 2018YFC1508001), the Fundamental Research Funds for the Central Universities (Grant Nos. B200204029) and the National Natural Science Foundation of China (Grant No. 51479061). The authors are grateful to the editor and anonymous reviewers for their invaluable comments and suggestions.
References (46)
Climate change and global water resources
Glob. Environ. Chang.
(1999)- et al.
Separating the impacts of climate change and human activities on streamflow: A review of methodologies and critical assumptions
J. Hydrol.
(2017) - et al.
Discrimination between climate and human-induced dryland degradation
J. Arid Environ.
(2004) - et al.
Variable selection using random forests
Pattern Recogn. Lett.
(2010) - et al.
Inundation simulation for urban drainage basin with storm sewer system
J. Hydrol.
(2000) - et al.
Trend analysis of rainfall in the headstreams of the Zambezi River Basin in Zambia
Phys. Chem. Earth
(2008) Climate change in Ireland from precipitation and streamflow observations
Adv. Water Resour.
(1999)- et al.
Investigation of urban climates and built environment relations by using machine learning
Urban Clim.
(2021) - et al.
Impacts of land use change and climate variability on hydrology in an agricultural catchment on the Loess Plateau of China
J. Hydrol.
(2009) - et al.
Detecting and attributing vegetation changes on China’s Loess Plateau
Agric. Forest Meteol.
(2017)
Hydrological analysis for water level projections in Taihu Lake, China
J. Flood Risk Manag.
Anthropogenic contributions dominate trends of vegetation cover change over the farming-pastoral ecotone of northern China
Ecol. Indic.
Impact of climate variability and human activity on streamflow decrease in the Miyun Reservoir catchment
J. Hydrol.
Integrating diverse methods to understand climate–land interactions in East Africa
Geoforum
Attribution of climate and human activities to vegetation change in China using machine learning techniques
Agric. Forest Meteol.
The long-term water level dynamics during urbanization in plain catchment in Yangtze River Delta
Agric. Water Manag.
The relative impact of urbanization and precipitation on long-term water level variations in the Yangtze River Delta
Sci. Total Environ.
Climate change impacts on extreme precipitation in Morocco
Glob. Planet. Chang.
Advances in separating effects of climate variability and human activity on stream discharge: An overview
Adv. Water Resour.
Effects of climate change and human activities on inflow into the Hoabinh Reservoir in the Red River basin
Procedia Environ. Sci.
Impacts of climate change and human activities on vegetation cover in hilly southern China
Ecol. Eng.
Unraveling the role of human activities and climate variability in water level changes in the Taihu Plain using artificial neural network
Water
Spatial and temporal variability of precipitation maxima during 1960–2005 in the Yangtze River basin and possible association with large-scale circulation
J. Hydrol.
Cited by (6)
Identifying changes in flood characteristics and their causes from an event-based perspective in the Central Taihu Basin
2023, Science of the Total EnvironmentWater quality variation and driving factors quantitatively evaluation of urban lakes during quick socioeconomic development
2023, Journal of Environmental ManagementChanges in flood characteristics and their driving mechanisms in the Taihu Plain region
2023, Dili Xuebao/Acta Geographica SinicaBibliometric analysis and global research trends of climate change and cities studies for 30 years (1990–2021)
2023, Environment, Development and Sustainability