1. Introduction
To mitigate the negative impact of urbanization on regional hydrology and environment, different initiatives or plans have been made worldwide, such as Low Impact Development (LID) or Best Management Practices (BMP) in the US [
1], Sponge City construction plan in China [
2,
3,
4], and Sustainable Urban Drainage System (SUDS) and Green Infrastructure in UK [
5,
6]. These plans all include building ‘Green Infrastructures’, or GIs [
7,
8] to decentralize stormwater management. GIs built for retaining and infiltrating stormwater runoff have been found very effective in runoff reduction and pollutants removal across various spatial scales [
9,
10,
11,
12,
13].
Existing studies mostly focused on physical performance of GIs, i.e., stormwater retention or pollutant removal efficiencies. The issue of financial responsibility in construction and operation of GI facilities has not been investigated thoroughly; though a few studies mentioned that financial concerns can be a major barrier for GI implementation [
5,
11,
14,
15,
16,
17,
18]. In order to reach full-scale GI implementation, we need to locate the responsible parties by examining the general pattern of urban development process. A city usually expands with addition of different Land Parcels (LPs), including industrial parks, transportation areas, cultural and educational facilities, business and residential areas. Owners or operators of different LPs need to understand that while they benefit from the land development, the change in land use will cause negative impacts on urban hydrology. They should take the mitigation responsibility and pay the additional cost for disposing the increased stormwater runoff, such as stormwater taxes [
19,
20]. Alternatively, they can install mitigation GIs to earn credits to avoid such cost or penalty [
18,
19]. Thus there is a great advantage to hold the LP owners as responsible for GI implementation [
5,
11,
15,
17]. However, city LPs vary considerably in land use, constructing GIs to reach the same storm runoff mitigation goal would result in different levels of difficulty in different LPs. Theoretically, it is possible to design GIs to make all LPs to meet the mitigation requirement [
3,
21,
22,
23], but this may not be economically feasible. The sponge city construction guideline of China set lower standards for some LPs with GI construction difficulty [
4]; such compromises, however, will have a negative effect on the overall compliance of building mitigation GIs. To avoid this shortcoming, we may look beyond an individual LP’s boundary, and look for suitable sites in neighboring LPs, in which GI construction may be easy so that extra stormwater retention capacity can be generated for credit trading [
18]. This capacity trading (CT) approach may create links among LPs so that overall storm runoff mitigation goals can be achieved through uneven construction of GIs.
Few studies have been conducted to explore planning and management of urban stormwater based on LPs, but the idea has been raised for some years. For example, Thurston (2006) [
17] and Thurston et al. (2003) [
18] cited the work by Pigou (1962) [
24] and argued that the optimal solution for combating pollution is to directly tax the parties that are responsible for. This has been materialized in several cases [
19,
20]. Meanwhile, the advances in monitoring technologies have made it possible to locate the responsible parties accurately [
25], and the modern technologies in managing spatial and temporal information can plan GIs on different or newly developed LPs at reasonable cost and speed [
5]. Unfortunately, few studies have been conducted on the effect of LPs, we postulate the following two reasons for the gap:
Existing GI projects are mainly supported by public funds on public domains, or on private properties with some government subsidies; many of these projects are research oriented, or for demonstration purposes. This to a great extent has made the economic concerns less pressing than it should be;
Many studies that were conducted at watershed scale or over large areas used mathematical models to evaluate hydrological and environmental impact of GIs; these modeling studies generally dealt with physical delineation lines, such as rivers, roads, and watershed boundaries, economical responsibilities were seldom considered [
3,
26].
Previous studies have investigated the economic advantages of building GIs over the conventional stormwater facilities [
17,
18], or the flexibility in implementation schemes [
10]. While they have encouraged more GIs implementations, the economic responsibility of GIs has not been clearly addressed. Considering the high upfront cost of GI implementation and the long-term maintenance requirement, economic analyses and optimization should be studied considering the constraints imposed by urban LP distributions [
5]. In this paper, we proposed a GI building capacity trading (CT) approach to optimize the construction cost in an urban development area; as a preliminary investigation on the feasibility and potential of CT, we demonstrated the procedures and effectiveness of CT under different trading conditions with a case study in the ancient city of Yangzhou, China.
3. Results
Under the baseline condition of GI implementation, the
RC of the study area was proposed to be reduced to the predevelopment level (0.5). The suggested
RC values for runoff retention facilities (
ψGI) that possess catchment area ratio (s) of 10:1 was 0.2 (Ministry of Resident and Contruction (MORC), 2014). For the baseline condition of converting 10% pervious surface into flow retention GI space, the
RCs for different LP types were computed as listed in
Table 2. Comparing with the target value (0.5), deficit LP types include commercial plazas (
RC = 0.70), large industrial areas (
RC = 0.58), and lower education schools (
RC = 0.59); surplus LP types include residential area (
RC = 0.36), higher education organizations (
RC = 0.22), and small industrial areas (
RC = 0.48).
As a preliminary experiment, we examined two CT scales: (1) neighboring trading, and (2) 20 m range trading scale, which is about the width of a medium size road. Following the procedures described in
Section 2.1, the number and area of deficit LPs were identified for the two trading scales and the results are listed in
Table 3. For better comparison, the overall CT was included as the fourth scenarios in
Table 3, and the four scenarios under consideration are as follows:
No capacity trading (No-CT),
Trading with neighboring LPs (Neighboring CT),
Trading with surplus LPs within 20 m range (20 m radius CT), and
Baseline condition, which is equivalent to overall trading (Overall CT).
Without CT, there were 139 LPs with deficit that covered 649 ha, which accounted for 39% and 38% of the total number and area of LPs in total (355 LPs over area of 1723 ha). With Neighboring CT, the number and area of deficit LPs were reduced to 97 and 558.4 ha, or about 30% reduction in number and 14% reduction in area comparing with the No-CT condition. When the trading scale extended to 20 m, the number of deficit LPs was further reduced to 78, and the deficit area was reduced to 478 ha; these is 40% reduction in number and 26% reduction in area of the deficit LPs.
Table 4 lists composition of the deficit and surplus LPs. For the No-CT condition, 100 deficit LPs (72%) are commercial area, 21 (15%) are industrial, and the rest 18 (13%) are lower education organizations. These ratios changed little under the neighboring trading scenario. Under the 20 m range trading scenario, the percentage of deficit commercial area lowered to 62%, while the percentage of industrial LPs increased from 15% to 23%.
5. Conclusions
Current GI constructions are mostly implemented by academies, public agencies, or private sectors with public cooperation; this situation has somehow hidden or shadowed the financial responsibility of construction. Thus, explicitly linking responsible parties from different LPs in urban development can be functional for large-scale implementation of GIs. Considering different land use in variable LPs in the urban landscape, reaching a stormwater control target across-the-boundary will encounter some difficulties or resistance in some LPs. Not to compromise the overall goal, the proposed CT approach may be used to coordinate the efforts across different property lines. Considering the special features of stormwater management, CT may be spatially restricted to adhere to the onsite treatment principle normally associated with GIs. The CT mechanism is to ensure that LPs creating retention capacity surplus will be rewarded, and LPs with building difficulties can meet the target in a more cost effective way. Otherwise, there will be no incentives for some LPs to produce extra stormwater retention capacities. To some extent, introduction of CT overcomes the shortcoming and yet preserves the advantage of using LP as the basic counting units.
In this paper, we only discussed the CT approach associated with the physical or land use differences, there may exist other factors that interfere with the trading processes, such as urban pipe network layout, surface elevation difference, construction cost, and so on. Findings from this study may serve as a bottom-line for further investigation on how to coordinates the multilateral efforts.