In recent years, the bibliometric literature review has been broadly performed in hydrology (Islam et al. 2021), ecology (X and Z 2021), regenerative medicine (Chen et al. 2012), and other fields. Using visualization software, one can directly, clearly, and vividly diagnose the dynamic evolution, development trend, research progress, and hotspots in the research field. CiteSpace (Chen 2004, 2017) is extensively utilized in analyzing keywords and citations of articles due to its powerful function, excellent visualization effects, and rich layouts. Keywords are direct descriptions of the content and theme of a study, and citations comprehensively display the study details and the development trend. In the keyword co-occurrence analysis, the frequency of keywords usually reveals the evolution of keyword hotspots in the field of research and the strength of individual keywords. In combination with burst terms, it can review the volatile hotspots in the research field and explore the potential applications. Keyword cluster analysis is to classify data into various groups according to the homogeneity between keywords to obtain high support directions of different categories. Co-citation analysis can be utilized to show the research content of key nodes, such that it is convenient for researchers to summarize current research clusters and analyze future research directions.
In the present scrutiny, a total of 2103 articles in the field of LID research are collected. Of which, 1053 articles written in English are retrieved from the Web of Science (WOS) database. The timeframe is between 2004 and September 2021. The keyword used in the literature search is low impact development. The literature type is limited to article. The CNKI database is also exploited for Chinese literature retrieval, the understudy timeframe is between 2009 and September 2021, the keyword used for search is “low impact development + LID”, and the literature type includes both journal publications and dissertations, and a total of 1050 articles are collected. Then, the CiteSpace5.6.R5 software is employed to analyze the quantity, subject categories, keywords, and co-citations of the included articles. This software is able to summarize the research status, progress, and hotspots, as well as the associated problems with those studies. Finally, research prospects are proposed and explained in some detail.
Bibliometric analysis of LID-based research articles
Based on the bibliometric analysis of the data derived from WOS and CNKI, the variation of the annual number of publications and the cumulative number has been demonstrated in Fig. 1. It can be seen that the cumulative numbers of articles written in Chinese and English are generally comparable. The LID research abroad emerged in 2004, and then the annual number of articles has steadily grown. The Chinese articles on the LID originated in 2009, and the number of articles rapidly increased after 2015. The concept of LID is originally proposed by the Prince George’s County Department of Environmental Resources in the US. Subsequently, on the basis of the LID concept, several other concepts are proposed globally, such as Australia's Water Sensitive Urban Design (WSUD), New Zealand's Low Impact Urban Design and Development (LIUDD) program, Sustainable Urban Drainage System (SUDS) in Europe and the Sponge Cities in China. In Graham et al. (Graham et al. 2004) published in 2004, the specific concept of "Low Impact Development" was employed for the first time, and the efficiency of the LID measures such as bioretention, green roofs, and permeable pavement was assessed through water balance simulation. In China, the construction of pilot sponge cities in 2015 can be regarded as a key time point in the LID research.
From the subject categories of the WOS literature (see Table 1), the largest categories of English articles are Environmental Sciences & Ecology, Environmental Sciences, and Water Resources. In these English-based studies, the runoff and pollution caused by heavy rainfall have been managed through source control such that the development area has a natural hydrological cycle as much as possible. Islam et al. (Islam et al. 2021) proposed the overall goals of the LID, which includes: (1) comprehensive control of the precipitation and water quality, (2) control of the runoff from the source as much as possible, (3) increasing urban hydrological processes, such as evaporation, infiltration, and storage, and (4) improving ecological benefits. In addition to the theoretically performed research, the application of LID in various fields, including engineering, civil engineering, and engineering and environment, has been examined by abroad researchers.
Table 1 The subject categories of the LID research based on the WOS database.
|
No.
|
Quantity
|
Centrality
|
Subject
|
|
1
|
689
|
0.35
|
Environmental Sciences & Ecology
|
|
2
|
637
|
0.11
|
Environmental Sciences
|
|
3
|
560
|
0.08
|
Water Resources
|
|
4
|
467
|
0.53
|
Engineering
|
|
5
|
232
|
0.16
|
Engineering & Civil
|
|
6
|
212
|
0.03
|
Engineering & Environmental
|
Different froms of the research abroad, the largest categories of Chinese-based studies include Construction Science & Engineering and Water Conservancy & Hydropower Engineering (see Table 2). This is mainly attributed to the fact that the rapid urbanization in China is the primary challenge for urban stormwater management. The increase in urban construction land has changed rivers outside the city into inland rivers, reducing stormwater storage space, increasing impervious areas, and aggravating urban waterlogging. Moreover, urban construction has a huge impact on the ecological environment and exerts great pressure on water security. With the construction of the pilot sponge cities in 2015, Chinese scholars have conducted extensive examinations on the difficulties and demands in the construction of sponge cities, which greatly expand the application field of the LID research.
Table 2 The subject categories of the LID research based on the CNKI database.
|
No.
|
Quantity
|
Centrality
|
Subject
|
|
1
|
919
|
61.27
|
Construction Science & Engineering
|
|
|
2
|
357
|
23.80
|
Water Conservancy & Hydropower Engineering
|
|
|
3
|
88
|
5.87
|
Environmental Science and Resource Utilization
|
|
|
4
|
72
|
4.80
|
Road and Water Transportation
|
|
|
5
|
8
|
0.53
|
Resource Science
|
|
|
6
|
4
|
0.27
|
Gardening
|
|
International LID research progress
Keyword co-occurrence analysis of research topics
The LID is a novel stormwater management method, and its whole evolution can be divided into the formation and development stages. According to the statistics of high-frequency keywords (see Table 3), the formation stage can be rationally considered in the time interval of 2004-2014. In this stage, the context of the LID research is finalized, and the impact on the hydrological cycle, model simulation, and practical application is discussed. The top keywords include Performance, System, as well as those representing crucial components of the hydrological cycle such as Runoff, Infiltration, and Water Quality. The SWMM is broadly utilized in the LID research due to its excellent hydrological and hydrodynamic simulation performance. In addition, Bioretention and Green Infrastructure are also repeatedly mentioned as potential application directions. The development stage of the LID started in 2015. In this stage, the focus is on the Optimization, Uncertainty, and Resilience plus to the Management Practice. The research content involves perfecting the LID research framework, as well as addressing new demands such as water security and water ecology brought about by global climate change.
Based on the burst terms in English-based articles (see Fig. 2), it can be found that Stormwater Management has been a research hotspot throughout the research period due to its central role in the LID-based research. The LID measures such as Porous Pavement and Bioretention emerged as short-term research hot topics. In recent years, the hotspots have shifted to Stormwater Management Model and Challenges due to global climate change.
Table 3 The high-frequency keywords of publications according to the WOS database
|
Stage
|
Keywords (Frequency/centrality)
|
|
Formation stage (2004-2014)
|
LID (503/0.04)；runoff (249/0.05)；performance (220/0.02)；
|
|
stormwater management (191/0.06)；system (153/0.09)；urbanization (140/0.07)；
|
|
bioretention (137/0.06)；green infrastructure (133/0.04)；water quality (113/0.04)；
|
|
model (107/0.03)；SWMM (105/0.03)；infiltration (79/0.08)
|
|
Development stage (2015-2021)
|
optimization (68/0.01)；management practice (63/0.01)；infrastructure (31/0.01)；
|
|
city (30/0.02)；urban runoff (28/0.01)；uncertainty (25/0.04)；
|
|
benefit (22/0.01)；resilience (17/0.01)；reduction (17/0.01)
|
Keyword cluster analysis
According to the keyword clustering (see Table 4) and the timeline view of clustering (see Fig. 3), it is detectable that the research on Bioretention (#0) is relatively independent, and the experimental analysis and water quality improvement (contaminant removal rate) have been the focus of the clustering study. In contrast, the other LID measures such as Green Roof (#2) and Green Infrastructure (#4) focus more on the model simulation and water reduction (e.g., total runoff, peak flow, and peak flood time). In addition, landscape ecological benefit is a performance indicator of all LID measures. Optimization (#3) and Climate Change (#7) are the focus of the current research, the clustering keywords such as Optimization, Uncertainty, Cost-effectiveness, Climate change, and Urbanization represent the hot topics and the leading edge of the current research. In the cluster, the cost has been considered as the objective function in some articles, and they are aimed to scrutinize the distribution strategy for reducing the total runoff, peak runoff, and pollutant load. With more and more in-depth investigations, some scholars have commenced considering the influences of externally uncertain conditions such as climate change, urbanization, and rainfall. Furthermore, the parameter optimization of the LID measures and multi-criteria methods (e.g., AHP, TOPSIS) has attracted much attention. The SWMM is the most common model exploited in the LID research. Kaykhosravi et al. (Kaykhosravi et al. 2018) conducted a comprehensive analysis of eleven LID models as a function of the model characteristics, hydrology, and hydraulic modules and found that the SWMM had the best applicability.
Table 4 The keyword clustering of the publications based on the WOS database
|
Cluster name
|
Size
|
Homogeneity
|
Research topic（log likelihood ratio /P value）
|
|
#0 bioretention
|
112
|
0.616
|
bioretention (48.76/0.001)；SWMM (31.51/0.001)；stormwater (30.66/0.0001)
|
|
#1 sponge city
|
106
|
0.457
|
sponge city (38.37/0.0001)；bioretention (18.038/0.001)；urban catchment (13.75/0.001)；water quality (13.68/0.001)；sustainablity (13.19/0.001)
|
|
#2 green roof
|
69
|
0.669
|
green roof (28.95/0.0001)；porous pavements (21.32/0.0001)；permeable pavement (21.04/0.0001)
|
|
#3 optimization
|
61
|
0.694
|
optimization (37.09/0.0001)；cost (17.07/0.001)；management practice (16.46/0.0001)；uncertainty (11.33/0.001)
|
|
#4 green infrastructure
|
57
|
0.697
|
green infrastructure (24.35/0.0001)；stormwater management (20.31/0.001)；urban hydrology (17.28/0.0001)
|
|
#5 low impace development
|
36
|
0.745
|
low impact development (32.02/0.0001)；SWMM model (28/0.0001)；LID (12.97/0.001)
|
|
#6 constant head test
|
34
|
0.829
|
constant head test (12/0.001)；pollution (12/0.001)；sensitivity analysis (10/0.005)
|
|
#7 climate change
|
33
|
0.784
|
climate change (35.15/0.0001)；urbanization (31.45/0.05)；land use (13.78/0.001)
|
|
#8 SWMM
|
28
|
0.873
|
SWMM (44.37/0.0001)；water quality (10.5/0.005)；conservation subdivision (10.17/0.005)
|
Co-citation clustering analysis
Table 5 The co-citation clustering of the LID research according to the WOS
|
Cluster
|
Size
|
Homogeneity
|
Average year
|
Research topic（log likelihood ratio /P value）
|
|
#0 multi-objective optimization
|
163
|
0.717
|
2017
|
multi-objective optimization (11.77/0.001)；optimization (10.98/0.001)；SWMM (8.97/0.005)；climate change (6.33/0.05)；sustainable urban drainage systems (6.02/0.05)；life cycle cost (5.77/0.05)
|
|
#1 hydrology
|
132
|
0.902
|
2008
|
hydrology (22.52/0.0001)；bioretention (14.78/0.001)；SWMM (12.47/0.001)；sponge city (12.23/0.001)；sustainable development (12.06/0.001)
|
|
#2 groundwater
|
85
|
0.781
|
2014
|
groundwater (9.36/0.005)；urban hydrology (9.08/0.005)；green infrastructure (12.47/0.001)；groundwater recharge (7.95/0.005)
|
|
#3 sponge city
|
71
|
0.796
|
2016
|
sponge city (25.75/0.0001)；bioretention (12.92/0.001)；urban sustainability (7.81/0.01)
|
|
#4 LID practices
|
57
|
0.875
|
2011
|
LID practices (6.77/0.01)
|
|
#5 green roof
|
52
|
0.897
|
2013
|
green roof (26.31/0.0001)；LID (12.2/0.001)；retention (10.79/0.005)
|
|
#6 first flush effect
|
46
|
0.95
|
2017
|
first flush effect (10.22/0.0005)；denitrification (10.22/0.005)；stormwater runoff (8.93/0.005)
|
|
#7 sustain
|
30
|
0.921
|
2011
|
sustain (6.93/0.01)；site selection (6.13/0.05)；LID-BMPs (6.13/0.05)
|
|
#8 conservation subdivision
|
26
|
0.979
|
2007
|
conservation subdivision (16.32/0.0001)；experimental auction (8.13/0.005)
|
|
#10 benefit cost ratios
|
21
|
0.952
|
2002
|
benefit cost ratios (10.43/0.005)；conservation (10.43/0.005)；water balance (7.67/0.01)
|
Based on the co-citation clustering (see Table 5) and the co-citation references clustering network (see Fig. 4), it can be seen that the LID research can be divided into the three major groups: optimization (#0 multi-objective optimization), principle (#1 hydrology, #2 groundwater, and #6 first flush effect), and application (#3 sponge city, #4 LID practices, #5 green roof). It can be seen from the clustering timeline (see Fig. 5) that the research on the principle of the LID originated in the field of hydrology. Since 2004, a large number of investigations have been carried out on the impact of the LID measures on the hydrological processes, which demonstrate that precipitation is the main factor that affects the hydrological process of the LID measures. Qin et al. (Qin et al. 2013) systematically examined extreme rainfall events with various precipitation, duration, and peak intensity, and analyzed the performance of the LID measures due to different precipitation characteristics. From 2012, urban hydrology and groundwater recharge have become the research hotspots. Palla and Gnecco (Palla and Gnecco 2015) conducted a simulation examination of the LID-based systems on the urban watershed scale and validated the effectiveness of the LID measures in various rainfall return periods. In the past few years, the research on the principle of the LID measures mostly focused on the initial scour effect, which is also the main reason of the urban nonpoint source pollution. Yang et al. (Yang and Chui 2018) proposed an appropriate classification standard for multi-objective stormwater management, which is employed to select suitable bioretention facilities considering multiple performance objectives (e.g., reduction of initial scour effect, and lessening of the total runoff and peak flow).
Table 6 The key publications from co-citation references clustering according to the WOS
|
Frequency
|
Year
|
Author
|
Title
|
|
122
|
2017
|
Eckart et al. (2017)
|
Performance and implementation of low impact development - A review
|
|
109
|
2015
|
Fletcher et al. (2015)
|
SUDS, LID, BMPs, WSUD and more – The evolution and application of terminology surrounding urban drainage
|
|
79
|
2015
|
Palla and Gnecco (2015)
|
Hydrologic modeling of Low Impact Development systems at the urban catchment scale
|
|
75
|
2016
|
Ahiablame and Shakya (2016)
|
Modeling flood reduction effects of low impact development at a watershed scale
|
|
75
|
2016
|
Chui et al. (2016)
|
Assessing cost-effectiveness of specific LID practice designs in response to large storm events
|
|
64
|
2013
|
Qin et al. (2013)
|
The effects of low impact development on urban flooding under different rainfall characteristics
|
|
61
|
2012
|
Ahiablame et al. (2012)
|
Effectiveness of low impact development practices: literature review and suggestions for future research
|
|
58
|
2015
|
Baek et al. (2015)
|
Optimizing low impact development (LID) for stormwater runoff treatment in urban area, Korea: Experimental and modeling approach
|
|
51
|
2015
|
Liu et al. (2015a)
|
Enhancing a rainfall-runoff model to assess the impacts of BMPs and LID practices on storm runoff
|
|
50
|
2015
|
Rosa et al. (2015)
|
Calibration and verification of SWMM for low impact development
|
|
49
|
2017
|
Kong et al. (2017)
|
Modeling stormwater management at the city district level in response to changes in land use and low impact development
|
|
47
|
2015
|
Jia et al. (2015)
|
LID-BMPs planning for urban runoff control and the case study in China
|
|
46
|
2019
|
Li et al. (2019)
|
Comprehensive performance evaluation of LID practices for the sponge city construction: A case study in Guangxi, China
|
|
45
|
2018
|
Eckart et al. (2018)
|
Multiobjective optimization of low impact development stormwater controls
|
|
44
|
2017
|
Xia et al. (2017)
|
Opportunities and challenges of the sponge city construction related to urban water issues in China
|
|
41
|
2015
|
Martin-Mikle et al. (2015)
|
Identifying priority sites for low impact development (LID) in a mixed-use watershed
|
|
40
|
2017
|
Mao et al. (2017)
|
Assessing the ecological benefits of aggregate LID-BMPs through modelling
|
|
40
|
2017
|
Li et al. (2017)
|
Sponge city construction in China: A survey of the challenges and opportunities
|
|
40
|
2018
|
Zhang and Chui (2018)
|
A comprehensive review of spatial allocation of LID-BMP-GI practices: Strategies and optimization tools
|
|
39
|
2015
|
Liu et al. (2015b)
|
Evaluating the effectiveness of management practices on hydrology and water quality at watershed scale with a rainfall-runoff model
|
From the research of LID applications points of view, scholars in other countries have examined green infrastructure and bioretention in the early stage. However, after 2015, the performed research works have been mainly focused on the construction of sponge cities in China. For instance, Xia et al. (Xia et al. 2017) and Li et al. (Li et al. 2017) summarized the opportunities and challenges facing the construction of sponge cities in China. LID optimization has been a hot topic since it was first mentioned in 2013. The relevant studies adopted the SWMM and optimization algorithm to solve the objective function (e.g., water quantity (Baek et al. 2015, Liu et al. 2015a), water quality (Liu et al. 2015b), and cost minimization (Chui et al. 2016), thereby determining the optimal deployment of the LID measures. Some other explorations also considered uncertainties in the input parameters such as precipitation, anti-seepage coefficients, accumulation and scour coefficients, as well as changes brought about by climate change (Sohn et al. 2019) and land-use type (Kong et al. 2017).
LID research progress in China
Keyword co-occurrence analysis of research topics
According to the high-frequency keywords listed in Table 7, the research contents of Chinese articles in the formation stage (i.e., the time interval of 2009-2014) are basically the same as those of English articles, mainly focusing on the nature of LID and sponge cities, as well as the applications of both the SWMM and SUSTAIN models. In the development stage (i.e., the time interval of 2015-present), different from research works abroad that have focused on global climate change, China’s research works have been mostly focused on solving the negative effects of urbanization, including an in-depth exploration of landscape design and sponge city construction such as rainfall gardens, LID facilities, landscaped gardens, and sponge campuses. It is should be noticed that at this stage, the total annual runoff control is proposed as the general target of the LID design in China, and the concept of sustainable development of sponge city construction is also defined.
Based on the burst terms displayed in Fig. 6, it can be found that present China’s LID research is mainly conducted in three branches: rainfall gardens, residential areas, and sponge campuses. The sponge campus is a small urban stormwater system and is a hot topic of current research, whose burst strength reaches 3.3. Additionally, a variety of approaches such as AHP and multi-objective optimization have been exploited to determine the optimal type, size, as well as quantity, and location of the LID measures.
Table 7 The high-frequency keywords of publications according to the CNKI database
|
Stage
|
Keywords（frequency/centrality）
|
|
Formation stage (2009-2014)
|
LID (756/1.21)；Sponge city (336/0.26)；SWMM mode (125/0.11)；
|
|
Stormwater management (57/0.03)；Rrainwater utilization (37/0.01)；Cost effectiveness(31/0.01)；
|
|
Urban waterlogging (29/0.01)；Runoff control (274/0.01)；Urbanization (24/0.01)；
|
|
Green roof(17/0.01)；Non-point source pollution (16/0.00)；SUSTAIN model(15/0.00)
|
|
Development stage (2015-2021)
|
Rainfall garden (35/0.01)；Landscape design (31/0.01)；LID facilities(22/0.04)；
|
|
Landscape garden (21/0.01)；Rainwater system (20/0.01)；total annual runoff control (15/0.00)；
|
|
Sponge campus (13/0.00)；Stormwater management model (12/0.00)；Sustainable development (11/0.00)
|
Keyword cluster analysis
From the keyword cluster analysis (see Table 8) and the timeline view of the clustering (see Fig. 7) standpoints, the demonstrated results reveal that the cluster is relatively simple. The keywords in the Grass trenches (#2), Rainwater Utilization (#3) and Sponge City (#4) clusters are mostly presenting the application areas. According to Fig. 7, the key nodes are concentrated in the early stage, reflecting the lack of in-depth theoretical research in China. In addition to the SWMM model simulation, the principles, design, performance, advantages and disadvantages, and costs of the LID facilities (e.g., bioretention ponds, grass trenches, concave green space, green roof, permeable pavement, rainwater tank, and roof truncation) are explored in some detail (Liu et al. 2021).
Table 8 The keyword clustering of publications based on the CNKI database
|
Cluster
|
size
|
Homogeneity
|
Research topic（log likelihood ratio /P value）
|
|
#0 LID
|
93
|
0.779
|
LID (71.82/0.0001)；Urban waterlogging ( 39.9/0.0001)；Sustainable development (18.18/0.0001)
|
|
#1 SWMM model
|
70
|
0.719
|
SWMM model (69.66/0.0001)；Sponge-type综合管廊 (45.16/0.0001)；SUSTAIN model (45.16/0.0001)
|
|
#2 Grass trenches
|
53
|
0.841
|
Grass trenches (45.54/0.0001)；Green roof( 35.47/0.0001)；permeable pavement (22.55/0.0001)
|
|
#3 Rainwater utilization
|
50
|
0.824
|
Rainwater utilization (43.52/0.0001)；LID technologies (35.14/0.0001)；Landscape design (26.85/0.0001)
|
|
#4 Sponge city
|
47
|
0.697
|
Sponge city (93.67/0.0001)；Benefit quantification ( 29.77/0.0001)；Optimization (29.77/0.0001)；Annual runoff control rate (23.15/0.0001)
|
Comparison of the LID research in China and abroad
By comparing the LID research in China and abroad, the following results are obtained:
(1) Research abroad focuses on the improvement and expansion of the LID framework, while China’s research emphasizes on the application of technologies associated with the sponge city construction. From the number of publications by subject category point of view, the research field with the largest number of publications in English articles is Environmental Science, whereas the top research field in China is Architectural Science. This issue is closely related to the severe water safety as well as water ecological problems in China. Compared to the relatively mature urban stormwater management systems in foreign countries, the rapid urbanization process in China has resulted in the lack of long-term planning for urban stormwater management systems. Since the proposal of sponge city construction in 2015, the number of related Chinese articles has grown exponentially. The conducted studies not only solve the problems of urban waterlogging, groundwater resource depletion, and urban nonpoint source pollution in China, but also provide references for the LID application in other countries.
(2) Commonly, the research abroad emphasizes on the impact caused by global climate change, whereas China’s research is mainly motivated by changes in land-use types. From the keyword clustering standpoints, research abroad basically focuses on the LID parameter optimization, modelling and improvement of the multi-objective optimization method, uncertainty due to climate and land-use changes, and disaster resistance, while China’s research emphasizes on problems encountered in sponge city construction. The advantage is that, in the context of urban waterlogging and nonpoint source pollution, such explorations can quickly and comprehensively analyze the problems and provide timely information for China’s sponge city construction, sustainable development, water safety, and water ecology. Nevertheless, there are also some apparent shortcomings, such as high homogeneity, and lack of innovation.
(3) The research abroad focuses on optimization of the LID parameters through experimental research, while China’s research emphasizes on the model simulation in order to determine the type, quantity, location, and combination of optimal LID measures. The bioretention represents the largest cluster of English articles, and Experimental Analysis and Water Quality Improvement are the focus of such investigations. However, few Chinese explorations have been carried out in the areas mentioned above. China’s research mainly employs the SWMM model to simulate the LID measures with obvious water reduction effects, such as concave green space and permeable pavement, subjected to various rainfall return periods.