A New Methodology for Updating Land Cover Maps in Rapidly Urbanizing Areas of Levying Stormwater Utility Fee
1
National Institute Environmental Research, Hwangyong-ro 42, Incheon 22689, Korea
2
Korea Institute of Ocean Science and Technology, Haeyang-ro 385, Busan 49111, Korea
*
Author to whom correspondence should be addressed.
Appl. Sci. 2022, 12(7), 3254; https://doi.org/10.3390/app12073254
Submission received: 21 January 2022
/
Revised: 18 March 2022
/
Accepted: 22 March 2022
/
Published: 23 March 2022
(This article belongs to the Section Environmental Sciences)
Abstract
:With a steady increase in impervious surfaces due to urbanization in Korea, there is a growing burden and an urgent need to fund better management of nonpoint sources of pollution and stormwater. A prerequisite for securing the necessary financial resources is the determination of basic data for the accurate calculation of impervious surfaces as the basis for estimating the costs of nonpoint source pollution control and billing of stormwater utility fees. This requires the extraction of landcover information in a Geographic Information System (GIS) environment and development of a landcover map that accurately delineates pervious surfaces within impervious surface areas. However, since landcover maps in Korea are generated and updated at irregular intervals, it is difficult to properly track land use changes. To address this problem, this study has developed a new method for the detailed updating of landcover maps in developing urban areas to facilitate the calculation of stormwater utility fees. Sejong City was selected as the study site because it has experienced large-scale land use changes due to the recent relocation of the national administrative capital and continuous urbanization. The methodology proposed in this study is based on various spatial data such as aerial photographs and digital topographic maps and follows four process steps: preprocessing, first and second updates of landcover information, and quality assurance. In a test of this method, a total of 19,049 reclassified items were generated in the first and second updates, affecting a total area of 26.49 km2 within the original landcover map. The accuracy of these updates reached 99.78%, considering the changed areas and rate of change. This study provides fundamental data for further application of a stormwater utility fee policy in Korea. However, further research is required to automate the generation of accurate pervious/impervious maps and develop pertinent guidelines so that individual municipal and provincial governments can generate and update their own pervious/impervious maps as a basis for calculating the impervious surfaces in their regions.
1. Introduction
Due to rising coverage by impervious ground surfaces resulting from rapid urbanization, Korea has experienced an increasing number of urban flash flooding incidents caused by stormwater runoff; these events transport pollutants from nonpoint sources directly into local streams or lakes [1]. Thus, impervious surfaces have direct impacts on stormwater runoff and water quality deterioration in urban watersheds. Despite the importance of impervious surfaces as an indicator of water quality at the watershed scale, Korea still lacks an effective stormwater management policy [2,3]. Advanced countries, such as the United States of America and Germany, levy stormwater utility fees to finance the costs of stormwater management caused by impervious surfaces [4,5,6,7,8,9]; Korea has also laid down certain standards for stormwater utility fees by land-surface class in the pertinent district unit planning guidelines [10]. At the implementation level, however, Korea lags behind other developed countries, being hampered by variable standards among governmental departments and a lack of interdepartmental coordination.
The development and implementation of a proper stormwater management plan based on the amount of impervious surface area should be preceded by the securing of resources to finance the management of nonpoint source pollution. One potential funding method is a stormwater utility fee, which can be determined by the ratio of impervious surface area to total parcel area according to the “polluter pays” principle [11]. Therefore, it is essential to prepare the basic data necessary for calculating the impervious surface area of a given parcel, i.e., area information by landcover class extracted in a Geographic Information System (GIS) environment and a landcover map clearly delineating pervious and impervious surfaces [12].
A landcover map is a spatial information database representing the shapes of ground objects in a given area after classification according to a set of scientific criteria and color-indexing of objects with similar characteristics. It reflects land surface features better than any other map type and hence is widely used for calculating the nonpoint source pollution load and the mapping of impervious surfaces.
Intensive research has been conducted on landcover map generation for the purpose of identifying land use changes and calculating the ground surface area by land use class, in order to solve the problem of nonpoint source management and pollutant load estimation. Lee et al. [13] generated landcover maps using low- and medium-resolution satellite imagery and various thematic maps, and calculated pollutant loads based on these areas by landcover class. Park et al. [14] reviewed the production process of existing medium-resolution landcover maps and improved the medium-resolution landcover map generation process using satellite imagery. Lee et al. [15] developed a method to update medium-resolution landcover maps based on various spatial data and used this to support the management of land-based pollutant sources. The estimation of impervious surface area requires a high-resolution landcover map to reflect current land use. At present, it is difficult to trace land use changes in Korea, given the relatively long and irregular four to five years’ map updating intervals [16]. Moreover, urbanization-induced dryland has been left barren or used in ways that deviate from the original land use class (such as artificial greenery); this has been aggravated by illegal land use changes and urban sprawl [17]. As a result, the actual land use status is often different from the cadastral land use or cannot be clearly defined, leading to inaccurate area estimations for impervious surfaces.
To reflect landcover changes due to rapid increases in the percent coverage of impervious surface area, it is necessary to update existing high-resolution landcover maps. With regard to spatial resolution, updating landcover at the subpixel level is also required to allow for the inclusion of smaller objects.
In consideration of these requirements, in this study we tested a method for the accurate and precise updating of Korean landcover maps with details based on various spatial databases in order to ensure accurate estimation of impervious surface areas; this would facilitate their accurate classification as the basis for nonpoint source management and stormwater utility fee collection in new urban areas. To this end, this study collected various spatial data and updated time-series items, corrected classification errors, and proposed a detailed landcover map update methodology that enables the detailed updating of pervious areas existing within impervious areas on an annual basis.
2. Materials and Methods
2.1. Study Site
Sejong City in west-central Korea was selected as the study site for the methodology proposed in this study (Figure 1). As the recently relocated national administrative capital, this region has an urgent need to manage impervious areas due to rapid, continuous urbanization and a high rate of landcover change. Sejong City occupies an area of 465.23 km2, 77% of the size of Seoul. As of March 2018, its population numbered about 290,000, and was projected to increase over time due to ongoing urbanization. The combination of continuing large-scale urbanization with a high landcover change rate made Sejong City a suitable site for testing this study’s methodology.
2.2. Definition and Collection of Available Spatial Data
In this study, the spatial data necessary for the annual updating of landcover maps were defined and collected according to the following criteria: (1) Annually updatable spatial data allowing temporal tracking of landcover changes; (2) Spatial data amenable to landcover classification according to the existing landcover map guidelines, given that updates are performed on existing landcover maps; (3) Spatial data from large-scale maps suitable for representing small-area landcover items matching the context of actual land use status; (4) Low-cost spatial data. Then the data meeting these criteria were collected. The 2015 high-resolution landcover map was collected by the Environmental Geographic Information Service of the Ministry of Environment as the baseline map for upgrading. A digital topographic map 2.0 (1:5000) of the National Geographic Information Institute was obtained to check the accurate boundaries of objects with impervious surfaces (such as buildings). In order to detect landcover changes by visual inspection and accurately check pervious surfaces existing within an impervious area, we collected ortho-aerial photos with 25 cm spatial resolution from the National Geographic Information Institute. Additionally, a portal site map service was used to check areas occluded by trees, shadows, and high buildings. The collected spatial data were provided free of charge with various map data and visualization information as an open service to the public, and are shown in Table 1.
2.3. Detailed Landcover Map Update Methodology
The GIS-based detailed urban area landcover map upgrade methodology is composed of four steps (Figure 2): preprocessing, first and second manual updating process of landcover information, and quality assurance.
These steps were grouped into stage 1 (coordinate conversion and first landcover update) and stage 2 (second landcover update and quality assurance) in order to facilitate the process depending on the characteristics of the landcover maps obtained, which comprised 7 low-level (First level, 1:50,000), 22 medium-level (Second level, 1:25,000), and 41 high-level (Third level, 1:5000) maps. Medium-level landcover maps were generated nationwide in 2007, but the generation of high-level landcover maps is still underway. Data from medium-level landcover maps, i.e., those lacking some details, were intended to go through both stages, while data from high-level landcover maps were intended to go only through stage 2.
High-level landcover maps were available for Sejong City, allowing direct stage 2 processing. However, given that the maps were generated in 2015, some updates and corrections were performed in stage 1. As of 2017, high-level landcover maps are available throughout the country, except for Jeollanam-do province in the south, and additional updating was performed in the northern and southern Han River areas.
2.3.1. Conversion Coordinates and Preprocessing
The coordinates of the spatial data collected then needed to be converted because they were generated by different organizations and in different modalities. All spatial data were placed in the same coordinate system and digitized after being subjected to preprocessing including numerical mapping, coordinate conversion, and boundary matching. High-level landcover map coordinate conversion was performed based on the Global Coordinate System ITRF 2000 TM central origin adopted by the standard national Digital topographic map 2.0.
2.3.2. First Update of Landcover Information
Currently, landcover maps produced in Korea follow the landcover map guidelines provided by the Ministry of Environment [18]. The landcover maps were produced by an on- screen digitizing method by visually classifying the boundaries and attributes of the landcover based on the image data of the relevant year [19]. The GIS-based high-resolution landcover maps used in this study were generated by the Ministry of Environment in 2014, which contained different landcover information for the areas changed after the onset of urbanization in 2015. In the first updating process, the classification items were updated to reflect the actual land use status in the areas estimated to have undergone temporal changes at the subpixel level of the high-resolution landcover map. In addition, the misclassifications on the existing high-resolution landcover maps were additionally updated. Since the areas originally classified as Used Area in low-resolution maps contained grass, more detail was necessary to reclassify such pervious surfaces within a larger impervious area as “other grass” and “other barren” to prevent them from being included in the calculation of impervious surface area. Details of such additional landcover information are defined in Table 2. For spatial object classification and updating, performed screen digitizing based on visual inspection of spatial data such as ortho-aerial photos and satellite images. In cases where object boundaries or attributes were blurred or ambiguous due to shadows, buildings, or trees, the portal site used their relationships with the adjacent objects and road views. When adding pervious surfaces reclassified as “other grass” or “other barren” within the Used Area, the building boundary was modified using the building layers on the numerical maps to prevent pervious areas from being included in buildings.
2.3.3. Second Update of Landcover Information
In stage 2, despite the use of various spatial data, a field survey was conducted to check areas occluded by shadows and buildings, and updated the corresponding items using the portal site’s road view. Also, a field survey was conducted on areas under construction and those with suspected changes. The accuracy of the landcover map updated by the first and second processes was tested by inspection of the updated items.
2.3.4. Landcover Misclassification Check and Quality Assurance
To ensure proper application of the proposed method to a new urban area nonpoint source pollution management program, its accuracy should be assured. Therefore, an error detection table was prepared using the quality assurance checklist proposed in the landcover map guidelines. Samples were extracted from the classified landcover items by random sampling per landcover class and their accuracy was rated by superimposing ortho-aerial photos and using the portal site map service. Accuracy testing was performed on both object boundaries (represented by polygons) and attributes, and either error was rated as misclassification. Errors were defined as “misclassification” (wrong attribute), “non-classification” (missing attribute data), “boundary correction” (wrong boundary contour), and “other” errors. For quality assurance (QA), the samples selected were 10% of the polygons, relative to the total number of objects. The error rate of the detailed classification by landcover map item is expressed by Equation (1), in which the overall error rate was obtained by averaging the error rates of all items. Items with error rates within 5% of the overall error rate were determined as the final products, and the first and second update processes were repeated until the cut-off error rate of 5% was reached.
- : Error rate
- : Number of misclassifications
- : Number of non-classifications
- : Number of boundary corrections
- : Number of other errors
- : Number of QA control samples
3. Results and Discussion
The landcover classification change rates after the first and second update processes are compared in Table 3.
3.1. Results of the First Detailed Landcover Map Update Process
In the first process, misclassified and reclassified items were updated according to the criteria and temporal changes, respectively. Then, detailed classification of the grassland and barren land areas existing within urban dryland (the broad classification) prevailing in new urban areas was updated. Figure 3 shows the results of the first and second updates for an apartment complex in Sejong City. The land-use classes in the existing landcover map consisted of “multi-family dwellings” (112); “commercial office and buildings” (131); “roads” (154); “other grass” (423); and “other barren” (623). The first update resulted in the addition of “cultural, sports, and recreational facilities” (141) and correction of “other grass” and “other barren” areas to roads and cultural, sports, and recreational facilities.
Furthermore, most items were changed in the high-resolution (large-scale) landcover map during the first update, most notably areas of the 14 detailed classifications for “Used Area”. Increased categories included “single-family dwellings” (0.14%); “multi-family dwellings” (0.13%); “industrial facilities” (0.13%); “commercial and office building” (0.15%); “cultural, sports, and recreational facilities” (0.09%); “railways” 0.01%; “roads” (1.67%); “educational and administrative facilities” (0.01%); and “other public facilities” (0.14%); these totaled 2.46%, with “roads” showing the highest increase rate. Unclassified shopping areas and parks within residential areas (e.g., apartment complexes) were added, along with boundary corrections. The post-2015 temporal land-use changes were not reflected in the high-resolution landcover maps of Sejong City, and the areas classified as “other barren” primarily arose during urbanization. The greenery areas within apartment complexes were corrected to “other grass” in the detailed classification of impervious surface areas.
3.2. Results of the Second Detailed Landcover Map Update Process
In the second update process, field surveys were performed to check areas that could not be classified based on aerial photos or maps, or were suspected of temporal changes. Figure 4 shows a typical correction based on a field survey in an area where the completion of a building site was not clearly marked on the map.
Through the field survey, most areas that were occluded by buildings and shadows on the map and could not be classified were changed to “roads” or “other grass,” and the existing parks within apartment complexes were classified as “other grass” after the completion of the landscaping work. As a result, all such items showed small changes in areas, such that “other grass” showed the largest change rate with 0.11%.
3.3. Results of the Update Accuracy Quality Assurance Control
The classification accuracy could be confirmed using various spatial data, and as a result, finding an accuracy rate of up to 97.9% (Table 4). This high accuracy is attributable to the visual inspection of spatial data with high spatial resolutions and the classification and boundary checks based on field surveys. However, a large number of occlusion areas, including trees/thickets and gardens in apartment complexes, resulted in lower accuracy of “artificial grassland” and “other barren land”, due to blurred boundaries compared with other classification items.
3.4. Accuracy Comparison between the First and Second Update Processes
A total of 19,049 polygons were newly generated in the landcover map update processes with respect to the existing landcover map, whereby the number of vertices increased by 762,973. In the first update process, boundary corrections were performed based on aerial photos with high spatial resolutions as well as detailed classification of pervious surfaces existing within areas classified as impervious surface area. As a result, the numbers of polygons and vertices increased by 17,726 and 732,787, respectively, relative to the existing landcover map. In the second update process, these increased further by 1323 and 30,186, respectively. Thus, in this study, through this two-tiered updating process, it was possible to generate a new map with very high accuracy. The degree of increase in the number of vertices indicates the degree of detail of the polygons (Table 5).
Comparison of the updated and existing landcover maps revealed class changes covering a total area of 26.49 km2. The accuracy of the first and second update results reached 99.78%, considering the changed area and rate of change. This suggests that errors can occur when calculating the impervious surface areas for nonpoint source management (Table 6).
Results demonstrated the accuracy of the proposed method, enabling the generation of updated landcover maps with substantially increased accuracy and precision in landcover classification and boundary delineation compared with the existing high-resolution (large-scale, 1:5000) landcover map in Korea.
4. Conclusions
Low Impact Development (LID) and Green Stormwater Infrastructure (GSI) are actively promoted in advanced countries in an effort to better manage stormwater runoff, mitigate the expansion of impervious surface areas, and minimize the adverse effects of impaired water circulation structures. In line with these trends, this study developed a method for detailed updating of high-resolution landcover map as well as high-accuracy map generation to facilitate the accurate calculation of stormwater utility fees as new financial resources capable of securing funding for LID and GSI projects. Given the inadequacy of existing Korean landcover maps for the accurate calculation of stormwater utility fees due to discrepancies between maps and actual land-use status, the upgrade method proposed here is expected to serve as a valuable basis for accurate calculation.
The updated high-resolution landcover maps were updated using various spatial data, and accurate classification of landcover map items and clear boundary delineation could be performed. By additionally classifying other grassland and barren land (such as greenery within apartment complexes) as impervious surfaces, landcover could be classified in a highly detailed and accurate manner.
In South Korea, notifications are absolutely mandatory when changing the use of land or completing constructions, and, among required notification items, address information should be included. If the map is updated by identifying landcover changes by the corresponding address through collection of address information and using up-to-date image data, then more effective editing, maintenance, and management of the map are possible. Such methods for updating maps would be applied in cases of artificial changes, as natural changes cannot be reflected. As urban areas mostly show artificial changes rather than natural changes, minimal effort may be required to maintain the newest traits and accuracy of the map.
However, the manual updating process based on spatial data is time consuming, making it necessary to develop an automated update algorithm to address this problem. Accurate and precise landcover maps generated and updated using the method proposed in this study will greatly contribute to facilitating systematic impervious surface management.
In this study, changes with time were not considered in the process of map generation; landcover changes with each hour, and change is more frequent in urban areas. Therefore, future methods should consider changes with time, as well as maintenance.
After the large-scale impervious surface map is generated, the ratio of impervious surface to individual land parcel can be easily calculated by overlaying the impervious surface map with a cadastral map in GIS. These statistics can be used as basic data for suggesting plans of management and policy regarding impervious surfaces. They can be used as a major factor in urban environment management, new development plans, and selecting priority management areas for impervious surfaces. This study serves as a guide for future efforts on generating and using large-scale impervious surface maps for effective and scientific management of the urban environment.
In addition, further research is needed to develop algorithms for automated pervious/impervious surface map generation based on the results of this study in order to enhance the efficiency of urban nonpoint source management.
Author Contributions
Conceptualization, J.Y. and C.L.; methodology, J.Y. and C.L.; validation, J.Y. and C.L.; formal analysis, J.Y.; investigation, J.Y.; resources, J.Y.; data curation, J.Y.; writing—original draft preparation, J.Y.; writing—review and editing, J.Y.; visualization, J.Y.; supervision, C.L.; project administration, J.Y.; funding acquisition, J.Y. All authors have read and agreed to the published version of the manuscript.
Funding
This research was funded by the Ministry of Trade, Industry and Energy (MOTIE, Korea).
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
Not applicable.
Acknowledgments
This work was supported by the Ministry of Trade, Industry andEnergy (MOTIE, Korea) under (20010447, Developed 6000 Dalton class UF membrane material and module). Their support is greatly appreciated.
Conflicts of Interest
The authors declare no conflict of interest.
References
- Kim, K.H.; Stephen, S.V.; Paul, M.H.; Peter, G.T.; Jeffrey, P. Urban Non-point-source Pollution Assessment Using a Geographical Information System. J. Environ. Manag. 1993, 9, 157–170. [Google Scholar] [CrossRef]
- Barnes, K.B.; Morgan, J.M., III; Roberge, M.C. Impervious Surfaces and the Quality of Natural and Built Environments; Department of Geography and Environmental Planning, Towson University: Baltimore, MD, USA, 2001; p. 21252. [Google Scholar]
- Booth, D.B.; Hartley, D.; Jackson, R. Forest Cover, Impervious-Surface Area, and The Mitigation Of Stormwater Impacts. JAWRA J. Am. Water Resour. Assoc. 2002, 38, 835–845. [Google Scholar] [CrossRef]
- Parikh, P.; Taylor, M.A.; Hoagland, T.; Thurston, H.; Shuster, W. Application of market mechanisms and incentives to reduce stormwater runoff: An integrated hydrologic, economic and legal approach. Environ. Sci. Policy 2005, 8, 133–144. [Google Scholar] [CrossRef]
- Roy, A.H.; Wenger, S.J.; Fletcher, T.D.; Walsh, C.J.; Ladson, A.R.; Shuster, W.D.; Thurston, H.W.; Brown, R.R. Impediments and Solutions to Sustainable, Watershed-Scale Urban Stormwater Management: Lessons from Australia and the United States. Environ. Manag. 2008, 42, 344–359. [Google Scholar] [CrossRef]
- Crockett, C.S. Parcel Based Billing for Stormwater; ASCE-Philadelphia Section; American Society of Civil Engineers (ASCE): Philadelphia, PA, USA, 2010. [Google Scholar]
- US EPA. Green Infrastructure Case Studies: Municipal Policies for Management Stormwater with Green Infrastructure, EPA-841-F-10-1004, EPA Office of Wetlands, Oceans and Watersheds. 2010. Available online: https://www.epa.gov/ (accessed on 20 January 2022).
- Bonnaffon, H. Local Stormwater Fees and Taxes: Results of the Sept, 2010 Survey; Chesapeake Bay and Water Resource Policy Committee (CBPC): Washington, DC, USA, 2011. [Google Scholar]
- Graham, S.; Schempp, A.; Troell, J. Regulating Nonpoint Source Water Pollution in a Federal Government: Four Case Studies. Int. J. Water Resour. Dev. 2011, 27, 53–69. [Google Scholar] [CrossRef]
- Lee, S.J.; Kim, H.S.; Kim, B.J.; Baek, J.S. The Core Keyword of Water Cycle Waterfront City. J. Water Policy Econ. 2017, 50, 43–49. [Google Scholar]
- Lee, C.; Kim, K.; Lee, H. GIS based optimal impervious surface map generation using various spatial data for urban nonpoint source management. J. Environ. Manag. 2018, 206, 587–601. [Google Scholar] [CrossRef]
- Kang, J.E.; Lee, M.J.; Koo, Y.S.; Cho, Y.H. Development and Application of Green Infrastructure Planning Framework for Improving Urban Water Cycle: Focused on Yeonje-Gu and Nam-Gu in Busan, Korea. J. Environ. Policy 2014, 13, 43–73. [Google Scholar]
- Lee, S.I.; Lee, C.S.; Choi, Y.S. Mapping the Distribution of Non-Point Source of Pollution Using Satellite Image Dataset and the Analysis of the Pollution Load Estimation Result. Korean Soc. Civ. Eng. 2003, 23, 719–726. [Google Scholar]
- Park, J.J.; Ku, J.Y.; Kim, B.S. Improvement of the Level-2 Landcover Map with Satellite Image. J. GIS Assoc. Korea 2007, 15, 67–80. [Google Scholar]
- Lee, C.Y.; Kim, K.H.; Kwak, G.H.; Oh, S.K. A Study on the Development of GIS-based Updating Methodology of Landcover Maps at Intermediate Level for Managing Nonpoint Source Pollutants. In Korean Water Congress 2014; Korean Society on Water Environment and Korean Society of Water and Wastewater: Goyang, Korea, 2014; pp. 66–67. [Google Scholar]
- Kwak, G.H.; Kim, K.H.; Lee, C.Y.; Oh, S.K. A Study on a GIS based Updating Methodology of Landcover Maps for the Enhancement of Utilization in the Total Maximum Daily Loads. J. Korean Soc. Water Environ. 2014, 30, 340–350. [Google Scholar] [CrossRef] [Green Version]
- Hong, S.E.; Yi, D.H.; Park, S.H. Land Category Non-coincidence Measurements Using High Resolution Satellite Image and Digital Topographic Maps. J. GIS Assoc. Korea 2004, 12, 43–56. [Google Scholar]
- Ministry of Environment (MOE). Guideline for producing Landcover Maps; Instruction No.1036; Ministry of Environment (MOE): Sejong, Korea, 2013.
- Environmental Geographic Information System (EGIS). 2022. Available online: https://egis.me.go.kr/ (accessed on 20 January 2022).
Figure 1.
Location of Sejong City, Korea.
Figure 2.
Flowchart of the proposed two-tiered detailed landcover map updating process.
Figure 3.
Example results of the first detailed landcover map update process, (a) Existing landcover map; (b) Landcover map after the first update.
Figure 3.
Example results of the first detailed landcover map update process, (a) Existing landcover map; (b) Landcover map after the first update.
Figure 4.
Example results of the second detailed landcover map update process, (a) Aerial photo; (b) After first update; (c) Field survey photo; (d) After second update.
Figure 4.
Example results of the second detailed landcover map update process, (a) Aerial photo; (b) After first update; (c) Field survey photo; (d) After second update.
Table 1.
Spatial data collection record.
| Thematic Map | Source | Updating Year and Cycle | Collection Route | File Format | Contents |
|---|---|---|---|---|---|
| High-level landcover map (1:5000) | Ministry of Environment (MOE) | Last updated in 2015 (4–5-year irregular interval) | Environmental Geographic Information Service | Shape file | A total of 108 map sheet |
| Digital topographic map 2.0 (1:5000) | National Geographic Information Institute | Updated weekly | National Geographic Information Institute Land Survey Division | Shape file | A total of 108 map sheet |
| Serial cadastral map | Local government (district office); Korea Land Information System | Updated daily | Portal Service of the Ministry of Land, Infrastructure and Transport | Shape file | Continuous map sheet |
| Ortho-aerial photos (GSD 25 cm) | National Geographic Information Institute | Aerial photos: 2014 Ortho-aerial photos: 2015 (every 2 years since 2010) | Spatial Information Service of the National Geographic Information Institute | GeoTiff | A total of 108 map sheet |
Table 2.
Definitions of the landcover items reclassified in detail.
| High-Level Classification | Objects of Detailed Classification | Changed Class | |
|---|---|---|---|
| Residential zone | Single family dwellings | Greenery within a single-family dwelling | Other grass |
| Multi-family dwellings | Greenery within an apartment complex | ||
| Industrial zone | Industrial facilities | Greenery and rest area within an industrial complex | |
| Commercial zone | Commercial office buildings | Greenery within a commercial zone | |
| Cultural, sports, and recreational zone | Cultural, sports, and recreational facilities | Sports facilities and recreational parks within an apartment complex; greenery within a sports park | |
| Playground and greenery within a stadium | |||
| Public facility zone | Educational and administrative facilities | Greenery within educational and administrative facilities | |
| Other public facilities | Greenery within other public facilities |
Table 3.
Comparison of change rates after the first and second update processes.
| High-Level Classification | Existing Landcover Map | Landcover Map after the First Update (Change Rate Relative to the Existing Landcover Map) | Landcover Map after the Second Update | ||
|---|---|---|---|---|---|
| Change Rate Relative to the Updated Landcover Map | Change Rate Relative to the Existing Landcover Map | ||||
| Proportion (%) | Proportion (%) | Proportion (%) | Proportion (%) | ||
| Used Area | Single family dwellings | 0.75 | 0.89 (+0.14) | 0.88 (−0.01) | 0.88 (+0.13) |
| Multi-family dwellings | 0.19 | 0.32 (+0.13) | 0.32 (-) | 0.32 (+0.13) | |
| Industrial facilities | 0.46 | 0.59 (+0.13) | 0.58 (−0.01) | 0.58 (+0.12) | |
| Commercial and office buildings | 0.61 | 0.76 (+0.15) | 0.77 (+0.01) | 0.77 (+0.16) | |
| Mixed areas | 0.00 | 0.00 (-) | 0.00 (-) | 0.00 (-) | |
| Cultural, sports, and recreational facilities | 0.11 | 0.20 (+0.09) | 0.20 (-) | 0.20 (+0.09) | |
| Airport | 0.00 | 0.00 (-) | 0.00 (-) | 0.00 (-) | |
| Harbor | 0.00 | 0.00 (-) | 0.00 (-) | 0.00 (-) | |
| Railways | 0.21 | 0.22 (+0.01) | 0.22 (-) | 0.22 (+0.01) | |
| Roads | 5.87 | 7.54 (+1.67) | 7.46 (−0.07) | 7.46 (+1.59) | |
| Environmental infrastructure | 0.02 | 0.03 (+0.01) | 0.03 (-) | 0.02 (-) | |
| Educational and administrative facilities | 0.12 | 0.13 (+0.01) | 0.13 (-) | 0.14 (+0.02) | |
| Other public facilities | 0.27 | 0.40 (+0.13) | 0.40 (-) | 0.40 (+0.13) | |
| Sum | 8.62 | 11.08 (+2.46) | 11.00 (−0.08) | 11.00 (+2.38) | |
| Agricultural Land | Classified rice paddies | 4.14 | 4.21 (+0.07) | 4.21 (-) | 4.21 (+0.07) |
| Unclassified rice paddies | 6.16 | 5.97 (−0.19) | 5.97 (-) | 5.97 (−0.19) | |
| Classified agricultural fields | 0.23 | 0.23 (-) | 0.23 (-) | 0.23 (-) | |
| Unclassified agricultural fields | 6.84 | 6.71 (−0.13) | 6.71 (-) | 6.71 (−0.13) | |
| Agricultural facilities | 0.59 | 0.59 (-) | 0.59 (-) | 0.59 (-) | |
| Orchards | 2.26 | 2.26 (-) | 2.26 (-) | 2.26 (-) | |
| Ranches/Aquafarms | 0.35 | 0.34 (−0.01) | 0.34 (-) | 0.34 (−0.01) | |
| Other cultivation facilities | 0.52 | 0.50 (−0.02) | 0.50 (-) | 0.50(−0.02) | |
| Sum | 21.08 | 20.81 (−0.27) | 20.81 (-) | 20.81 (−0.27) | |
| Forest | Broad-leaved deciduous forests | 25.46 | 25.19 (−0.27) | 25.19 (-) | 25.19 (−0.27) |
| Coniferous forests | 12.70 | 12.66 (−0.04) | 12.66 (-) | 12.66 (−0.04) | |
| Mixed forests | 5.31 | 5.15 (−0.16) | 5.15 (-) | 5.15 (−0.16) | |
| Sum | 43.47 | 43.00 (−0.47) | 43.00 (-) | 43.00 (−0.47) | |
| Grass | Natural grassland | 0.03 | 0.03 (-) | 0.03 (-) | 0.03 (-) |
| Golf courses | 0.17 | 0.17 (-) | 0.17 (-) | 0.17 (-) | |
| Cemeteries | 2.11 | 2.09 (−0.02) | 2.09 (−0.02) | 2.09 (−0.02) | |
| Other grass | 12.70 | 12.92 (+0.22) | 13.03 (+0.11) | 13.03 (+0.33) | |
| Sum | 15.01 | 15.21 (+0.20) | 15.32 (+0.11) | 15.32 (+0.31) | |
| Wet Land | Inland wetland | 1.87 | 1.87 (-) | 1.87 (-) | 1.87 (-) |
| Tidal flats | 0.00 | 0.00 (-) | 0.00 (-) | 0.00 (-) | |
| Salt pools | 0.00 | 0.00 (-) | 0.00 (-) | 0.00 (-) | |
| Sum | 1.87 | 1.87 (-) | 1.87 (-) | 1.87 (-) | |
| Barren | Beaches | 0.00 | 0.00 (-) | 0.00 (-) | 0.00 (-) |
| River banks | 0.17 | 0.17 (-) | 0.17 (-) | 0.17 (-) | |
| Rock faces, rocks | 0.02 | 0.02 (-) | 0.02 (-) | 0.02 (-) | |
| Mining areas | 0.03 | 0.03 (-) | 0.03 (-) | 0.03 (-) | |
| Playgrounds | 0.07 | 0.07 (-) | 0.07 (-) | 0.07 (-) | |
| Other barren | 7.40 | 5.49 (−1.91) | 5.46 (−0.03) | 5.46 (−1.94) | |
| Sum | 7.69 | 5.78 (−1.91) | 5.75 (−0.03) | 5.75 (−1.94) | |
| Water | Rivers | 1.92 | 1.92 (-) | 1.92 (-) | 1.92 (-) |
| Lakes | 0.33 | 0.33 (-) | 0.33 (-) | 0.33 (-) | |
| Coastal waters | 0.00 | 0.00 (-) | 0.00 (-) | 0.00 (-) | |
| Sum | 2.25 | 2.25 (-) | 2.25 (-) | 2.25 (-) | |
| Total | 100 | 100 | 100 | 100 | |
Table 4.
Error detection table for the detailed landcover map update results.
| Quality Assurance (QA) Items | Total Number of Polygons (QA-Controlled Polygons) | Type of Errors | Total Error (n) | Error Rate (%) | ||||
|---|---|---|---|---|---|---|---|---|
| Medium-Level Classification | High-Level Classification | ➀ Mis-Classification | ➁ Non-Classifi-Cation | ➂ Boun-Dary Correction | ➃ Other Errors | |||
| Residential zone | Single family dwellings | 8489 (2598) | 22 | - | 10 | - | 32 | 0.38 |
| Multi-family dwellings | 19,471 (4246) | 16 | - | 8 | - | 24 | 0.12 | |
| Industrial zone | Industrial facilities | 2258 | 64 | - | 1 | - | 65 | 2.88 |
| Commercial zone | Commercial and office buildings | 3689 (486) | 70 | - | 3 | - | 73 | 1.98 |
| Mixed areas | 2876 | 22 | - | 5 | - | 27 | 0.94 | |
| Cultural, sports, and recreational zone | Cultural, sports, and recreational facilities | 4986 | 58 | - | 11 | - | 69 | 1.38 |
| Traffic zones | Airport | - | - | - | - | - | - | - |
| Harbor | - | - | - | - | - | - | - | |
| Railways | 2246 | - | - | 3 | - | 3 | 0.13 | |
| Roads | 8781 (1986) | - | - | 12 | - | 12 | 0.14 | |
| Others | 159 | - | - | - | - | - | - | |
| Public facilities | Environmental infrastructure | 56 | - | - | - | - | - | - |
| Educational and administrative facilities | 536 | - | - | - | - | - | - | |
| Other public facilities | 1056 (18) | 23 | - | 38 | - | 61 | 5.78 | |
| Paddies | Classified rice paddies | 5865 | 15 | - | 4 | - | 19 | 0.32 |
| Unclassified rice paddies | 7853 | 20 | - | 6 | - | 26 | 0.33 | |
| Fields | Classified agricultural fields | 5496 | 16 | - | 4 | - | 20 | 0.36 |
| Unclassified agricultural fields | 7784 | 18 | - | 2 | - | 20 | 0.26 | |
| Agricultural facilities | Agricultural facilities | 2532 | - | - | 50 | - | 50 | 1.97 |
| Orchards | Orchards | - | - | - | - | - | - | - |
| Other cultivation facilities | Ranches/ Aquafarms | - | - | - | - | - | - | - |
| Other cultivation facilities | 1267 | - | - | - | - | - | - | |
| Forests | Broad-leaved deciduous forests | 9411 | - | - | 16 | - | 16 | 0.17 |
| Coniferous forests | 9547 | - | - | 10 | - | 10 | 0.10 | |
| Mixed forests | 9473 | - | - | 20 | - | 20 | 0.21 | |
| Natural grass | Natural grassland | 157 | 3 | - | 10 | - | 13 | 8.28 |
| Artificial grass | Golf courses | 24 | - | - | 4 | - | 4 | 16.67 |
| Cemeteries | 153 | - | - | - | - | - | - | |
| Other grass | 10,231 (3148) | 40 | - | 416 | - | 456 | 4.46 | |
| Inland wetland | Inland wetland | 247 | - | - | - | - | - | - |
| Coastal wetland | Tidal flats | - | - | - | - | - | - | - |
| Salt pools | - | - | - | - | - | - | - | |
| Natural barren | Beaches | - | - | - | - | - | - | - |
| River banks | - | - | - | - | - | - | - | |
| Rock faces, rocks | - | - | - | - | - | - | - | |
| Artificial barren | Mining areas | - | - | - | - | - | - | - |
| Playgrounds | 146 | - | - | 5 | - | 5 | 3.42 | |
| Other barren | 7952 (876) | 38 | 2 | 375 | - | 415 | 5.22 | |
| Inland waters | Rivers | 45 | - | - | 2 | - | 2 | 4.44 |
| Lakes | - | - | - | - | - | - | - | |
| Coastal waters | Coastal waters | - | - | - | - | - | - | - |
| Total | 132,786 (13,358) | 425 | 2 | 1015 | - | 1442 | 2.06 | |
Table 5.
Comparison of the number of polygons and vertices.
| Existing Landcover Map | First Updated Landcover Map (Added Items) | Second Updated Landcover Map (Added Items) | |
|---|---|---|---|
| Polygons (n) | 112,529 | 130,255 (+17,726) | 131,578 (+1323) |
| Vertices (n) | 7,217,490 | 7,950,277 (+732,787) | 7,980,463 (+30,186) |
Table 6.
Comparison of changed areas between the first and second update processes.
| Existing vs. First Updated Landcover Maps | First vs. Second Updated Landcover Maps | |
|---|---|---|
| Changed area (km2) | 25.45 | 1.04 |
| Change rate (%) | 5.47 | 0.22 |
| Matching rate (%) | 94.53 | 99.78 |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
MDPI and ACS Style
Yoo, J.; Lee, C. A New Methodology for Updating Land Cover Maps in Rapidly Urbanizing Areas of Levying Stormwater Utility Fee. Appl. Sci. 2022, 12, 3254. https://doi.org/10.3390/app12073254
AMA Style
Yoo J, Lee C. A New Methodology for Updating Land Cover Maps in Rapidly Urbanizing Areas of Levying Stormwater Utility Fee. Applied Sciences. 2022; 12(7):3254. https://doi.org/10.3390/app12073254
Chicago/Turabian StyleYoo, Jaehyun, and Cholyoung Lee. 2022. "A New Methodology for Updating Land Cover Maps in Rapidly Urbanizing Areas of Levying Stormwater Utility Fee" Applied Sciences 12, no. 7: 3254. https://doi.org/10.3390/app12073254
Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.
