Zone II • Versions EN2
Abstract: Impervious surface is one of the most basic components of a city which has a significant impact on urban ecological environment and regional development. Remote sensing has been widely used in impervious surface research in recent years due to its fast, large-scale, multi-scale and reproducible ground observation advantages. Taking Sanya City as the research area, this study selected and collected the Landsat series remote sensing image data of 2004, 2008, 2011, 2013 and 2015 in this area. A method combining BCI index and BI index is proposed to extract the impervious surface information in the study area. Based on this method, the impervious surface distribution maps of the five phases of Sanya City in 2004, 2008, 2011, 2013 and 2015 were obtained. Precision verification using Google Earth images which have high accuracy. It was concluded that this method is of strong operability, simple and easy to be applied to the research process while maintaining accuracy and efficiency, the highest precision is over 90%. This dataset can provide reference for the rational planning of “sponge city” in Hainan Province or other provinces and cities.
Keywords: impervious surface; remote sensing; ecological environment; sponge city
|Title||A dataset of impervious surface in Sanya City from 2004 to 2015|
|Data corresponding author||Meng Qingyan (firstname.lastname@example.org)|
|Data authors||Wang Xiankai, Meng Qingyan, Liu Ying, Yao Zhixin|
|Time range||From 2004 to 2015|
|Spatial resolution||30 m|
|Data volume||25.1 MB|
|Data service system||<http://www.sciencedb.cn/dataSet/handle/721>|
|Sources of funding||Natural Science Foundation of Hainan Province (417218); Hainan Province Major Science and Technology Plan (ZDKJ2016021); Science and Technology Support Program of Sichuan Province (2016JZ0027).|
|Dataset composition||This dataset includes Sanya City’s 2004–2015 Phase 5 impervious surface distribution data, all data is stored in a compressed file. The compressed file contains 5 Tif files, which correspond to the impervious distribution of Sanya City in 2004, 2008, 2011, 2013 and 2015. The total amount of data is 25.1 MB.|
Impervious surfaces are land-covering surfaces that prevent water from seeping into the soil, including roads, parking lots, plazas, roofs, and non-pervious surfaces in urban buildings. The increase of impervious surface area means that the increase of regional construction land and the acceleration of urbanization process is an indicator to measure regional development and has important reference value for rational planning of urban construction. Impervious surface can not only characterize the degree of urbanization, but also can be used to evaluate the urban ecological environment. In 2010, about 50% of the world's population lived in cities, and China's urban population reached 47%. Along with the acceleration of urbanization, the number of cities in China, especially large and medium-sized cities, has increased rapidly, and the scale of urban construction land has been expanding. The city has experienced rapid development and high-intensity development, resulting in a sharp increase in the surface area of impervious land. As of 2008, the average ratio of impervious surface in Chinese cities was about 66%. With the increase of impervious surface, environmental pollution has also intensified, and the urban landscape pattern has undergone dramatic changes. Therefore, it is of great significance to carry out research on impervious surface.
This study takes Sanya City as the research area. Sanya City is also known as Lucheng. It is located at the southernmost tip of Hainan Island. It is located between 18°09′34′′–18°37′27′′ north latitude and 108°56′30′′–109°48′28′′ east longitude. There are four districts in Tianya, Jiyang and Yazhou. The whole area is surrounded by mountains in the north and the sea in the south. The terrain gradually slopes from north to south, forming a narrow and long polygon. Sanya City has played a significant role in promoting the construction and development of Hainan Province and has played an important role in Hainan's development strategy. In recent years, Sanya City has entered the stage of rapid development of urbanization. The rapid development of Sanya City has led to a dramatic expansion of urban land use and dramatic changes in land cover types. According to the “13th Five-Year Plan”, optimizing urban master planning and land use planning has become one of the important tasks for the development of Sanya City. Therefore, as an important monitoring indicator for urban planning and land use planning, urban impervious surface condition is particularly important for urban ecological environment construction. Accordingly, based on Landsat data, the datasets uses the combination of Biophysical Composition Index (BCI) and Bare Soil Index (BI) to extract the impervious surface information of Sanya City, and obtains the impervious surface distribution map of the study area. Compared with other datasets of the same kind, the data source interval of this dataset is roughly the same, so that the research results can better reflect the data gradual process and update the time phase; the method of extracting impervious surface information using BCI index and BI index can ensure the accuracy and reliability of the data; this data set can be shared publicly for other applications. This paper studies the characteristics of urbanization construction in Sanya City by analyzing the changing law of impervious surface, and then provides a basis for scientifically and reasonably delimiting the development boundary and protection area of the city, maximizing the protection of the original ecological system, and providing a basis for the construction of "sponge city".
In this study, Landsat remote sensing images were taken as the data source to preprocess the image data, and the impervious surfaces information extraction method combining BCI index and BI index was constructed. Then, the impervious surfaces information extraction was carried out for a long time series in the research area, and finally the impervious surface distribution dataset of sanya in recent 10 years was generated.
2.1 Data collection
This dataset is based on the Landsat basic data of 2004, 2008, 2011, 2013 and 2015 at intervals of 2-4 years from 2004 to 2015. The data is mainly obtained from geospatial data cloud (http://www.gscloud.cn/), with less cloud cover and better quality. The specific experimental data is shown in Table 1 below, and the remote sensing image of Sanya City is shown in Fig 2.
2.2 Data processing
2.2.1 Data preprocessing
Remote sensing images perceive target information through sensors mounted on satellites, aircraft and other platforms without direct contact with the target, and characterize the spatial position and spectral characteristics of the target object through the changes of pixel spatial information and brightness value. Due to the lack of direct contact, the target information is affected by many factors in the transmission process. There are some geometric distortion and radiation distortion in the remote sensing image obtained by the sensors. In order to solve the above problems, the image data are pre-processed such as geometric correction, radiation correction and image registration in order to reflect the spectral information and geometric location information of the surface as truly as possible.
2.2.2 Impervious surface information extraction method
(1) BCI index
The BCI index was first proposed by Deng et al. It is based on the brightness (TC1), greenness (TC2) and humidity (TC3) components generated by TC transformation. For BCI index, the impervious surface is positively correlated with it and the gray value is greater than 0, the gray value of vegetation and other land cover is less than 0, and it is negatively correlated with vegetation coverage, and the gray value of soil is close to 0., so that the three components can be separated. Considering the influence of water body, before calculating BCI index, the water of the pretreated Landsat data were removed by Modification of Normalized Difference Water Index (MNDWI), and then BCI index was calculated.
Where: Green is the spectral reflectance of the green band; MIR is the spectral reflectance of the mid-infrared band. After the Landsat data is processed by removing water, etc., the BCI index is calculated by the following formula:
In the formula, H is the normalized TC1 component with high reflectance, L is the normalized TC3 component with low reflectance, and V is the normalized TC2 component with vegetation. The formulas for calculating the three factors are as follows:
Where: TCi (i = 1, 2, 3) is the first three TC components; TCimin and TCimax are the minimum and maximum values of the ith TC component, respectively.
(2) BI index
BI index was put forward by Rikimaru et al. in 2002. It can extract soil information effectively according to the characteristics that the reflectance of soil in red and short-wave infrared bands is higher than that of other surface cover types, while in mid-infrared band and blue-light band it is lower than that of other surface cover types. The BI index ranges from 0 to 200, and the bigger the BI value, the greater the probability of soil. In this way, soil information can be extracted by setting thresholds. In this way, soil information can be extracted by setting thresholds. To a large extent, soil information can be effectively eliminated. Compared with the Normalized Difference Soil Index (NDSI), the BI index performs better. The formula for calculating BI value is as follows:
Where: BLUE, RED, NIR and SWIR1 correspond to the spectral reflectance of blue, red, near infrared and short-wave infrared bands in remote sensing images respectively.
(3) BCI index combined with BI index to extract impervious information
The method of extracting impervious surface based on BCI has some advantages over other indexes, but it is still difficult to distinguish between soil and impervious surface. Some soils which have high BCI values are easily confused with impervious surface. In order to improve the accuracy of impervious surface extraction, after obtaining the impervious surface extraction result by the BCI method, this study introduced BI index, using BI index to extract soil information in the study area, and then effectively eliminate soil information to a large extent. Using the combination of BCI index and BI index to extract impervious surface can obtain more accurate results.
2.2.3 Long-term sequence impervious surface information extraction in the study area
On the premise of a given study area, determine the optimal data source type, spatial resolution and information extraction method of the city impervious surface information extraction input. The impervious surface of the city in the study area was analyzed in a long time series to obtain the impermeable water surface information of the city in the corresponding years. After processing the Landsat image data of the research area for nearly 10 years, a time series of impervious surface distribution products were formed.
2.2.4 Generating dataset
The extraction results of impervious surface information in the study area were made into data sets. This dataset can intuitively present the distribution of impervious surface and the spatio-temporal evolution characteristics of the research area, thus reflecting the direction of urban development and construction of the research area. The results are helpful to promote the monitoring of urban environmental changes in demonstration areas and provide technical process support for urban planning and environmental protection.
Based on remote sensing image data, the impervious surface information of Sanya City from 2004 to 2015 was extracted, and the distribution of impervious surface in different phases was generated. According to the data obtained, the distribution and evolution characteristics of impervious surface in Sanya can be obtained effectively. Fig.3 shows that a-e is the distribution map of impervious surface in 2004, 2008, 2011, 2013 and 2015, and fig.4 is the expansion map of impervious surface in Sanya from 2004 to 2015.
Based on the distribution maps of impervious surface of Sanya and the expansion maps of impervious surface of Sanya from 2004 to 2015, it can be concluded that the impervious surface of Sanya mainly concentrates in the south of Yazhou District, southeast of Tianya District, Jiyang District and Haitang District. Because the northern part of Sanya City is mountainous area, the distribution of impervious surface is relatively small. From the annual change of impervious surface, the impervious area of Sanya increased gradually from 2004 to 2015, and the degree of aggregation of impervious surface increased. The increase of coastal area was more obvious in regional change. And through the analysis of Sanya's impervious surface expansion, it can be predicted that Sanya's impervious surface will continue to expand along the coastal areas in the future, and gradually extend to the Inland areas. Internal filling and extension will be carried out at the same time, gradually forming the connecting trend among cities, and the urbanization level of Sanya will continue to improve.
In order to ensure the final generation of high-quality data sets of impervious surface distribution in the study area, the accuracy of the method of extracting impervious surface information combined with BCI index and BI index was verified. Based on the Landsat 8 remote sensing image of Sanya City in 2013 (Fig. 5a), the typical area with bare soil distribution was selected as the experimental area (Fig. 5b). Two methods, BCI index and BCI index combined with BI index, were used to extract the impervious surface information, and the difference in extraction accuracy between the two methods was compared. When BCI index is used, the results of impervious surface extracted are shown in Fig. 5c. When two indexes are combined, the results are shown in Fig. 5d. The white part is the pervious area and the black part is the impervious area. By comparing the extraction results, it can be concluded that the BCI index is misclassified as impervious soil information in Fig. 5C alone can be effectively eliminated to a great extent by the method of combining two indexes as the Fig. 5d describe, and the extraction accuracy of impervious surface is improved. The method is simple and easy to operate, and can be applied to large-scale impervious surface extraction.
Therefore, the BCI index and BI index are combined to extract the impervious surface of Sanya in 2004, 2008, 2011, 2013 and 2015, and the validation sample points are selected uniformly in the whole region. They are divided into two types: impervious surface and pervious surface. The extraction accuracy of impervious surface is quantitatively tested by Google Earth high-precision image accuracy verification. The results show that the method has good research effect, and the highest accuracy is over 90%. The test results are shown in Table 2.
An important manifestation of urbanization is the increase of impervious surface distribution ratio. Impervious surface is an important index to measure the level of urban development. The change of the distribution of impervious surface within a city has an important impact on the urban ecological environment. For large-scale cities, using high cost-effective medium spatial resolution images to obtain the distribution of impervious surface is the current international research hotspot. Taking Landsat image as data source and Sanya City as research area, this study produced five data sets of impervious surface distribution from 2004 to 2015, verified the applicability of index combination method to extract impervious information on a larger spatial and temporal scale, and provided basic data for ecological environment assessment, construction of "sponge city" and sustainable development of Sanya City.
Li D R, Luo H, Shao Z F. Application and prospect of remote sensing technology in impervious layer extraction [J]. Journal of Wuhan University (Information Science Edition), 2016 (05): 569-577.
Liu Y, Meng Q Y, et al. Estimation method of impervious surface coverage based on feature optimization and support vector machine[J]. Geography and Geo-Information Science, 2018 (1): 24-31.
Cheng X, S Z F , L J C, et al. Estimation of impervious surface coverage using hybrid spectral decomposition and SVM[J]. Journal of Remote Sensing, 2011 (06): 1228-1241.
Population Reference Bureau. 2010 World Population Data Sheet[R]. Washington D C: National Press Club, 2010.
Zhai W H, Liu J Y, Zhang Z X, et al. Remote sensing monitoring and spatiotemporal analysis of artificially constructed impervious surface in China in the early 21st century[J]. Chinese Science Bulletin, 2013, 58(56): 465–478.
BIERWAGEN B G, THEOBALD D M, PYKE C R, et al. National housing and impervious surface scenarios for integrated climate impact assessments [J]. PNAS, 107 (49): 20887-20892.
Ren F Y, Wei S C. Evaluation of Economical and Intensive Use of Regional Construction Land——Taking Sanya City as an Example[J]. Shanghai Land Resources, 2016, 37(3): 35-39.
DENG C, WU C. BCI: A biophysical composition index for remote sensing of urban environments [J]. Remote Sensing of Environment, 2012, 127: 247-259.
RIKIMARU A, ROY P S, MIYATAKE S. Tropical forest cover density mapping[J]. Tropical Ecology, 2002(43): 39-47.
Deng S B, Chen Q J, Du H J. ENVI remote sensing image processing method (second version)[M]. Beijing: Higher Education Press,2014:60-61.
XU H. Modification of normalised difference water index (NDWI) to enhance open water features in remotely sensed imagery[J]. INTERNATIONAL JOURNAL OF REMOTE SENSING, 2006,27(14):3025-3033.
RNDSI:A ratio normalized difference soil index for remote sensing of urban/suburban environments[J]. International Journal of Applied Earth Observation and Geoinformation, 2015, 39:40-48.
Wang X, Meng Q, Liu Y & Yao Z. A dataset of impervious surface in Sanya City from 2004 to 2015. Science Data Bank, DOI: 10.11922/sciencedb.721 (2019).
How to cite this article
Wang X, Meng Q, Liu Y & Yao Z. A dataset of impervious surface in Sanya City from 2004 to 2015. China Scientific Data 4(2019). DOI: 10.11922/csdata.2018.0067.zh