Mapping Green space, Roads &amp; Built-up Areas with Optical Remote Sensing and Polarimetric SAR (A Novel Approach)

doi:10.21203/rs.3.rs-4850131/v1

Download PDF

Research Article

Mapping Green space, Roads & Built-up Areas with Optical Remote Sensing and Polarimetric SAR (A Novel Approach)

https://doi.org/10.21203/rs.3.rs-4850131/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Recent progress in remote sensing and Geographical Information System (GIS) has revolutionized the research studies on urban space. Satellites that image on daily basis and cloud platforms that increase mathematical modelling precision and speed have given rise to extraction of high-quality data. In this study, we propose a novel approach in extracting information on green space, buildings, and roads in Ankara and Eskişehir cities in Türkiye. In this approach, optical and SAR images are utilized. Modelling is also fulfilled in google earth engine cloud platform using machine learning algorithm. We show how optical and SAR images with varying indexes may lead to a Land use/Land cover map with the highest overall accuracy (98.94 for Ankara and 93.97 for Eskişehir). Additionally, techniques offered in this study can help to extract other classes other than the present study ones. Land use/Land cover map is the basis of many studies and can benefit urban management, planning, urban policy making, protection and renovation, and environmental sustainment.

remote sensing in urban studies

google earth engine

machine learning algorithms

land use/land cover

urban form

Urbanization and quick increase of population have negatively affected urban space. Each city has its own challenges and evaluation of existing conditions always take priority when implementing a program. Technology helps to constantly monitor cities and shortens the process of data extraction for critical conditions. Being cognizant of the conditions, examining the factors, and trying to plan and resolve them make it possible to offer better urban services, increase urban life quality, and make urban space livable. Recent advances in remote sensing and GIS have revolutionized the access to data, regional modelling, and spatial analysis. New satellites can take images of cities daily. Moreover, cloud systems that have been devised to process these data are able to analyze and examine data quickly and accurately.

Using remote sensing satellites is one the methods of data collection for urban studies. Optical and SAR satellites have been employed in different studies. Landsat, Sentinel-2, and PlanetScope satellites are optical and Sentinel-1 satellite is a SAR satellite. Landsat-series satellite has been monitoring the earth for 50 years (since 1972). Its images are the richest archive of the earth and one of its purposes is monitoring the regional changes over time. The information from this satellite are the basis of many studies regarding forest monitoring (Townshend et al., 2012; Banskota et al., 2014), change detection (Awty-Carroll et al., 2019; Hemati et al., 2021), water quality (Peterson, Sagan and Sloan, 2020; Al-Shaibah et al., 2021), land surface temperature (Balew and Korme, 2020; Ermida et al., 2020) and mapping vegetation (Schwieder et al., 2016; Peterson, Sagan and Sloan, 2020). The images from Landsat 8 satellite have also been used in urban studies such as urban heat island (Elmes et al., 2020), ecological evaluation of urban heat island (Dissanayake, Kurugama and Ruwanthi, 2020), quantification of carbon sequestration by urban forest (Uniyal et al., 2022), extracting urban impervious surfaces (Deliry, Avdan and Avdan, 2021),urban green infrastructure health assessment (Chang et al., 2021), and producing land use/land cover (LULC) maps (Nasiri et al., 2022; Theres and Selvakumar, 2022).

Satellites differ in terms of spatial, spectral, radiometric, and temporal resolution. Landsat 8 satellite has 16-day temporal resolution, 30-meter spatial resolution, and 15-meter panchromatic band. Compared with Landsat 8 satellite, the Sentinel-2 satellite has 10-day temporal resolution that becomes 5 days because of using two repeat cycle satellites. This satellite also has bands with 10, 20, and 60-meter resolutions. Data produced by this satellite are typically used in studies regarding climate change, land monitoring, emergency, management, and security (Sentinel-2 User Handbook, 2015; Gibson et al., 2020; Mngadi, Odindi and Mutanga, 2021). Some studies show that Sentinel-2 satellite performs better than Landsat 8 satellite (J. Wang et al., 2020; Q. Wang et al., 2020; Ghayour et al., 2021). Landsat 8 satellite and Sentinel-2 satellite images can be combined (Mandanici and Bitelli, 2016). Using the images of these two satellites can remarkably increase the phenological variation and, consequently, precision of maps (Nasiri et al., 2022; Pouya, Aghlmand and Karsli, 2022). PlanetScope satellite has better temporal resolution than Landsat 8 and Sentinel-2 satellites. This satellite, which consists of 130 small satellites, takes images of the earth surface every day and may even take images of some regions more than once. What distinct this satellite from Landsat 8 and Sentinel-2 satellites is the 3-meter resolution of its images (Planet Team, 2018). Its images can play a critical role in urban studies. However, few studies have been conducted in this regard. sentinel − 1 satellite offers ordinary and systematic data to monitor sea and land, emergency reaction, climate change, and security. Two underlying limitations of optical satellites are lack of imaging at night and bad weather conditions (existence of cloud) (Sherpa and Shirzaei, 2022). Sentinel-1 satellite does not have these restrictions. Combining its images with those of Sentinel-2 and Landsat 8 satellites has been utilized in different studies on mapping of crop types and crop sequences (Blickensdörfer et al., 2022), eliminating leaf area index and aboveground biomass of grazing pastures (Wang et al., 2019), surface moisture and vegetation cover analysis (Urban et al., 2018), urban change detection (Benedetti, Picchiani and Del Frate, 2018; Hafner, 2022), and river water mapping (Liu et al., 2022) that provided higher quality results. As yet Sentinel-1 satellite images alone have not been used in producing land use maps. However, many studies have combined them with optical satellites images such as Landsat 8 and Sentinel-2 that has led to increased accuracy of the maps (Brandmeier et al., 2022; De Luca et al., 2022).

Regarding monitoring urban growth, extensive studies have been carried out using remote sensing satellites images and GIS. Most of these studies have been based on Landsat series images with 30-meter resolution and deal with urban growth. In these studies, archived images are received to examine changes and growth `extent of urban space to propose strategies for planning to associative officials (Dhanaraj and Angadi, 2020; Roy and Kasemi, 2021; Wang, Murayama and Morimoto, 2021; Elhamdouni, Arioua and Karaoui, 2022). Most of the studies have generally dealt with urban growth issue because of restrictions such as lack of access to images with high spatial resolution and platforms with high processing power and have not examined the growth of roads, paths, and urban green space separately. The results obtained with these images are based on their spatial resolution. Thus, small streets and paths cannot be mapped. Additionally, green spaces that have a smaller area than spatial resolution do not appear in the results. For example, in Landsat 8 satellite, the area of each pixel is 900 square meters. Thus, the three main classes of green space, roads, and buildings are not clear well in these images. In Fig. 1, which shows one neighborhood of Eskişehir city in Türkiye, some parts of urban space that has low-area buildings, roads, and green space can be observed. In Fig. 1, three images of PlanetScope satellite with 3-meter resolution (a), Sentinel-2 satellite with 10-meter resolution (b), and Landsat 8 satellite with 30-meter resolution (c) can be seen. The main image position in this figure has been shown by red. This figure clearly indicates the spatial resolution of these three satellites. In the image with 3-meter resolution, different elements such as green space, roads, and buildings can be easily seen; but in image b, separation of classes is hard and in image c, is impossible.

Using new satellites with better spatial, spectral, radiometric, and temporal resolution has helped to gain more information on urban space to produce more detailed LULC maps. Although images with higher resolution produce more details of urban space, their spatial resolution is low, and it is necessary to combine the images of satellites to reach better results. Combining images can also increase the bulk of data and process of them. All the process needs to be done in cloud platforms. Google earth engine (GEE) cloud platform can be referred to as one of the most important progress in GIS and remote sensing area that not only offer GIS data but also makes it possible to do different analyses on them. GEE data do not require pre-processing. Moreover, not only data of its database can be used but also those of other sources can be uploaded. This, in turn, makes it possible to analyze a great deal of data in a short time simultaneously. Another advantage of GEE platform is the existence of strong libraries for processing of satellite images. These libraries offer important machine learning algorithms for classification and production of LULC maps.

In this research, we aimed to produce a 3-meter LULC map and then extract data from green space, roads, and buildings. The study was conducted by using optical images (PlanetScope and Sentinel-2 ), varying indexes estimated from the bands of these two satellites, and also images of SAR satellite in GEE cloud platform with machine learning algorithm. The study was done on two cities in Türkiye, that is, Ankara and Eskişehir to examine data and offer the best combination of data to achieve the study purposes.

Research Area

Ankara and Eskişehir, as two cities of Türkiye, were selected as the research area of this study. They are located in the central Anatolian region of Türkiye (Fig. 2). According to Koppen-Geiger climate taxonomy, the studied area has Mediterranean summer warm and dry weather with cool winter. Ankara was selected as the capital of Türkiye in 1923. Its population was 2,747,325 in 1980 and is 5,747,325 now. In terms of population, Ankara is the second city of Türkiye and 57th of the world. Eskişehir population was 543,802 in 1980 and 898,369 now. It is ranked as the 25th city of Türkiye in terms of population. Ankara and Eskişehir are different from each other in social changes, environment, economic conditions, technology, planning, urban housing and policy, and transportation. This has caused them to have thoroughly different urban forms and spatial features.

Methodology

Figure 3 shows the process of the present study. It was conducted in two stages: data collection and criterion, and data processing. A detailed description of the two stages are provided in the following.

Data collection and creation

In this study, Sentinel-2, Sentinel-1, and PlanetScope images were utilized. Image selection was done based on the normalized difference vegetation index (NDVI). GEE provides the best possibility for extracting NDVI time series so that the months when this index is minimum, and maximum would be identified. In the study, the dataset of Ankara city was set based on the month when NDVI index was maximum (August) and that of Eskişehir was based on the month when the NDVI index was minimum (January).

Sentinel-2 and senbtinel-1 images are accessible in pre-processed forms in GEE platform. Images of PlanetScope were also downloaded in Tiff file and were then uploaded in GEE. The image of Sentinel-2 was set with 10 bands and resolutions of 10 and 20 meters in the platform. Further, NDVI, the normalized difference water index (NDWI), the Pigment Specific Simple Ratio (PSSR), Urban Index (Ui) and the Normalized Difference Built-up Index (NDBI) indexes from Sentinel-2 were estimated in GEE and were set in dataset. NDWI, NDVI, the Green Normalized Difference Vegetation Index (GNDVI) and PSSR indexes from PlanetScope satellite and two images from Sentinel-1 were also opted with ascending (A) and descending (D) directions. Each of the images had vertical transmit/horizontal receive (VH) and vertical transmit/vertical receive (VV) bands that were put in the dataset as well. Thus, 27 bands were included in the dataset 15 of which related to Sentinel-2, 8 ones related to PlanetScope, and 4 ones were related to Sentinel-1 (Fig. 3 and Table 1).

Table 1

Research dataset
No.	Satellite	Combination	The number of bands
1	PlanetScope	B4, B3, B2 & B1	4
2	PlanetScope	NDVI, NDWI, PSSR & GNDVI	4
3	Sentinel-2	B12, B11, B8A, B8, B7, B6, B5, B4, B3 & B2	10
4	Sentinel-2	PSSR, NDVI, NDWI, NDBI & Ui	5
5	Sentinel-1	VVA & VHA	2
6	Sentinel-1	VVD & VHD	2
Total bands			27

Data processing

After preparing dataset, the process of producing LULC map via Support Vector Machines Library (LibSVM) (Chang and Lin, 2011) was fulfilled. For Ankara, classes Build-up 1, Build-up 2, Vegetation-1, Vegetation-2, Road, Bare land and water; and for Eskişehir classes Build-up 1, Build-up 2, Vegetation-1, Road, Bare land and water were considered. Vegetation-1 pertained to trees and vegetation-2 pertained to bushes. Build-up 1 was related to buildings with Roof tiles and Build-up 2 was related to buildings with non-Roof tile. After selecting classes, 400 points were chosen for each class using GEE. Then, the selected points were checked based on RGB picture of PlanetScope, Sentinel-2, and indexes of NDBI, NDWI, AND NDVI. 70 percent of the points was considered for modelling (training points) and 30 percent (test points) for examining the accuracy of the model (overall Kappa coefficient & accuracy).

The result of classification is production of urban LULC map. This map was estimated with 3-meter resolution with Table 1 dataset that included images of the three above-mentioned satellite with different indexes from PlanetScope and Sentinel-2 images. It was done in GEE platform using LibSVM method. The accuracy and efficiency of the classification can be calculated through overall accuracy and Kappa coefficient. It turned out to be 98.94 and 98.34 for Ankara and 93.97 and 93.1 for Eskişehir, which are high accuracy.

Table 2

Results of classification by using LibSVM method
Ankara		Eskişehir
Overall accuracy	Kappa coefficient	Overall accuracy	Kappa coefficient
98.94	98.34	93.97	93.1

Figure 4 is Ankara’s LULC. In this figure, two areas (a and b) have been highlighted. As the figure shows, the two classes are clearly distinct from each other. Figure 5 also displays the results of extraction of each of the classes of urban green space, roads, and buildings from LULC map. As the Figs. 4 and 5 show, in some streets of Ankara, tall trees have caused the roads class not to be clearly visible.

Figures 6 and 7 also uncover the result of classification and production of LULC map for Eskişehir. They indicate that each of the classes are clearly distinct from others. The dataset of this city was based on the month when NDVI was minimum (January). The green space is the least in this month but the accuracy rate of ways (roads and streets) was high. Figure 7 clearly indicates the effect of using this method on accuracy of classification. Additionally, the urban green space in this figure is the least. Thus, the farmlands around the city are also visible in this figure.

Urban planning typically refers to activities that provide development plans for monitoring the use of urban spaces with the purpose of living conditions development and welfare of dwellers in urban areas. Urban planning plays a key role in urban dwelling and economic development of cities. It is taken as an important tool in decision making by local governments. Urban planning needs to be supported by different sciences. One of them is related to data extraction science. In order to pinpoint the current problems and needs and have a comprehensive understanding of the interactive components and their impacts, urban planners must take a wide range of information to analyze with some techniques and tools. They often obtain this information by field operations, document analysis, and typographic maps, which are costly, cumbersome, and time consuming.

In recent years, remote sensing technology has considerably been used to obtain information on urban areas. Fundamental urban planning needs to include urban plan strategies, city development and expansion purposes, standards and criteria for general urban building and land uses, comprehensive transportation system, and green space system. Comprehensive urban planning has been a complicated periodic program that needs a wide range of information that is prepared partly by GIS and remote sensing. Some merits of remote sensing and GIS systems include wide cover of satellite images, multi-time and multi-range imaging, image receive at different weather conditions, being economical, and easy access. Data obtained from GIS and remote sensing can be used to examine and analyze cities natural conditions, resource distribution, roads network, and land use changes.

Green space, roads, and buildings are among important elements of cities. Recent progress in remote sensing and GIS has made it possible to monitor them. New processing systems such as GEE and remote sensing satellite images have made it possible to analyze them freely and quickly. There are, however, some limitations such as technology-related constraints and weather condition-related restrictions.

In urban space, we need images that have high resolution, because spatial resolution is a determinant in mapping of green space, roads, and buildings. For example, in a image with 3-meter resolution, the green space that has 9 square meter area can form a complete one pixel in satellite image and this one pixel can be a tree or a small green space in a yard, school, or street. Thus, the images of PlanetScope satellite were used in the present study as they have 3-meter resolution with four bands. Although spatial resolution is high, spectral resolution is low and it can be stated that four bands of Green, Blue, Red, and NIR are not enough for classification. Therefore, four indexes of NDVI, NDWI, PSSR, and GNDVI were also estimated and added to the dataset. Increasing dataset leads to increasing machine learning algorithms efficiency that had remarkable classification accuracy by adding bands of Sentinel-2 along with PSSR, NDVI, NDWI, NDBI, and Ui indexes and VH and VI bands of Sentinel-1 images (Table 2). Adding Sentinel-1 bands amended errors such as tree shadows and improved classification accuracy in vegetation classes (trees and lawn). Another positive impact of Sentinel-1 was related to extraction of data from roads. Using the ascending and descending bands helped to increase their related classification (Figs. 4 and 5).

One of the errors that generally happen in classification is erroneous categorization of phenomena. For example, a point that has been identified as building is categorized in roads classification or a point selected as road may be placed in bare land classification. This error is the least in Ankara classification because the scope of urban green space, roads, and buildings is considerable which, in turn, increases the accuracy of classification and mapping. In small cities such as Eskişehir with smaller green space, roads, and buildings, it is more likely to see errors. Combining other data can minimize these errors. In fact, combing data and putting SAR data and different indexes can increase the features of each pixel of image. GEE cloud platform is a revolution in remote sensing and GIS system. Almost all remote sensing data are present in it in pre-processed form. Additionally, if specific data does not exist in it, it can be uploaded from other sources. Another advantage of GEE is the possibility of parallel analysis. This causes the calling of data, calculation of indexes, classification, and accuracy assessment to be fulfilled in less than 60 seconds. In this study, supervised classification offered by GEE was used. They are free to access. In addition, they can be revised as well. The other merit of this platform is the existence of aerial photos in base map and possibility of simultaneous show of varying satellites and indexes that eases the selection of training points for supervised classification in the shortest possible time.

Conflict of interest disclosure

The authors declare no conflict of interests.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author Contribution

This research is related to the doctoral thesis of Majid Aghlmand, who was the second and third author of the supervisors.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Al-Shaibah, B. et al. (2021) ‘Modeling water quality parameters using landsat multispectral images: a case study of Erlong Lake, Northeast China’, Remote Sensing, 13(9), p. 1603. doi: https://doi.org/10.3390/rs13091603.
Awty-Carroll, K. et al. (2019) ‘Using continuous change detection and classification of landsat data to investigate long-term mangrove dynamics in the Sundarbans region’, Remote Sensing, 11(23), p. 2833. doi: https://doi.org/10.3390/rs11232833.
Balew, A. and Korme, T. (2020) ‘Monitoring land surface temperature in Bahir Dar city and its surrounding using Landsat images’, The Egyptian Journal of Remote Sensing and Space Science, 23(3), pp. 371–386. doi: https://doi.org/10.1016/j.ejrs.2020.02.001.
Banskota, A. et al. (2014) ‘Forest monitoring using Landsat time series data: A review’, Canadian Journal of Remote Sensing, 40(5), pp. 362–384. doi: https://doi.org/10.1080/07038992.2014.987376.
Benedetti, A., Picchiani, M. and Del Frate, F. (2018) ‘Sentinel-1 and sentinel-2 data fusion for urban change detection’, in IGARSS 2018–2018 IEEE International Geoscience and Remote Sensing Symposium. IEEE, pp. 1962–1965.
Blickensdörfer, L. et al. (2022) ‘Mapping of crop types and crop sequences with combined time series of Sentinel-1, Sentinel-2 and Landsat 8 data for Germany’, Remote sensing of environment, 269, p. 112831.
Brandmeier, M. et al. (2022) Synergetic use of Sentinel-1 and Sentinel-2 data for large-scale Land Use/Land Cover Mapping. Copernicus Meetings.
Chang, C.-C. and Lin, C.-J. (2011) ‘LIBSVM: A library for support vector machines’, ACM transactions on intelligent systems and technology (TIST), 2(3), pp. 1–27. Available at: https://doi.org/10.1145/1961189.1961199.
Chang, M.-E. et al. (2021) ‘Urban green infrastructure health assessment, based on landsat 8 remote sensing and entropy landscape metrics’, European Journal of Remote Sensing, 54(1), pp. 417–430. doi: https://doi.org/10.1080/22797254.2021.1948357.
Deliry, S. I., Avdan, Z. Y. and Avdan, U. (2021) ‘Extracting urban impervious surfaces from Sentinel-2 and Landsat-8 satellite data for urban planning and environmental management’, Environmental Science and Pollution Research, 28(6), pp. 6572–6586. doi: https://doi.org/10.1007/s11356-020-11007-4.
Dhanaraj, K. and Angadi, D. P. (2020) ‘Land use land cover mapping and monitoring urban growth using remote sensing and GIS techniques in Mangaluru, India’, GeoJournal, pp. 1–27.
Dissanayake, K., Kurugama, K. and Ruwanthi, C. (2020) ‘Ecological evaluation of urban heat island effect in colombo city, Sri Lanka based on landsat 8 satellite data’, in 2020 Moratuwa Engineering Research Conference (MERCon). IEEE, pp. 531–536. doi: https://doi.org/10.1109/MERCon50084.2020.9185277.
Elhamdouni, D., Arioua, A. and Karaoui, I. (2022) ‘Monitoring urban expansion using SVM classification approach in Khenifra city (Morocco)’, Modeling Earth Systems and Environment, 8(1), pp. 293–298.
Elmes, A. et al. (2020) ‘Mapping spatiotemporal variability of the urban heat island across an urban gradient in Worcester, Massachusetts using in-situ Thermochrons and Landsat-8 Thermal Infrared Sensor (TIRS) data’, GIScience & Remote Sensing, 57(7), pp. 845–864. doi: https://doi.org/10.1080/15481603.2020.1818950.
Ermida, S. L. et al. (2020) ‘Google earth engine open-source code for land surface temperature estimation from the landsat series’, Remote Sensing, 12(9), p. 1471. doi: https://doi.org/10.3390/rs12091471.
Ghayour, L. et al. (2021) ‘Performance evaluation of sentinel-2 and landsat 8 OLI data for land cover/use classification using a comparison between machine learning algorithms’, Remote Sensing, 13(7), p. 1349.
Gibson, R. et al. (2020) ‘A remote sensing approach to mapping fire severity in south-eastern Australia using sentinel 2 and random forest’, Remote Sensing of Environment, 240, p. 111702.
Hafner, S. (2022) ‘Multi-Modal Deep Learning with Sentinel-1 and Sentinel-2 Data for Urban Mapping and Change Detection’. KTH Royal Institute of Technology.
Hemati, M. et al. (2021) ‘A systematic review of landsat data for change detection applications: 50 years of monitoring the earth’, Remote sensing, 13(15), p. 2869. doi: https://doi.org/10.3390/rs13152869.
Liu, Q. et al. (2022) ‘Fusing Landsat-8, Sentinel-1, and Sentinel-2 Data for River Water Mapping Using Multi-dimensional Weighted Fusion Method’, IEEE Transactions on Geoscience and Remote Sensing.
De Luca, G. et al. (2022) ‘Integrated use of Sentinel-1 and Sentinel-2 data and open-source machine learning algorithms for land cover mapping in a Mediterranean region’, European Journal of Remote Sensing, 55(1), pp. 52–70.
Mandanici, E. and Bitelli, G. (2016) ‘Preliminary comparison of sentinel-2 and landsat 8 imagery for a combined use’, Remote Sensing, 8(12), p. 1014.
Mngadi, M., Odindi, J. and Mutanga, O. (2021) ‘The utility of Sentinel-2 spectral data in quantifying above-ground carbon stock in an urban reforested landscape’, Remote Sensing, 13(21), p. 4281.
Nasiri, V. et al. (2022) ‘Land Use and Land Cover Mapping Using Sentinel-2, Landsat-8 Satellite Images, and Google Earth Engine: A Comparison of Two Composition Methods’, Remote Sensing, 14(9), p. 1977. doi: https://doi.org/10.3390/rs14091977.
Peterson, K. T., Sagan, V. and Sloan, J. J. (2020) ‘Deep learning-based water quality estimation and anomaly detection using Landsat-8/Sentinel-2 virtual constellation and cloud computing’, GIScience & Remote Sensing, 57(4), pp. 510–525. doi: https://doi.org/10.1080/15481603.2020.1738061.
Planet Team (2018) ‘Planet Application Program Interface: In Space for Life on Earth’. San Francisco, CA. Available at: https://api.planet.com.
Pouya, S., Aghlmand, M. and Karsli, F. (2022) ‘Looking into the green roof scenario to mitigate flash flood effects in Mamak, Turkey, via classifying images of Sentinel-1, 2, and PlanetScope satellites with LibSVM algorithm in Google Earth Engine cloud platform’.
Roy, B. and Kasemi, N. (2021) ‘Monitoring urban growth dynamics using remote sensing and GIS techniques of Raiganj Urban Agglomeration, India’, The Egyptian Journal of Remote Sensing and Space Science, 24(2), pp. 221–230.
Schwieder, M. et al. (2016) ‘Mapping Brazilian savanna vegetation gradients with Landsat time series’, International journal of applied earth observation and geoinformation, 52, pp. 361–370. doi: https://doi.org/10.1016/j.jag.2016.06.019.
Sentinel-2 User Handbook (2015). European Space Agency. Available at: https://sentinel.esa.int/documents/247904/685211/Sentinel-2_User_Handbook.
Sherpa, S. F. and Shirzaei, M. (2022) ‘Country-wide flood exposure analysis using Sentinel‐1 synthetic aperture radar data: Case study of 2019 Iran flood’, Journal of Flood Risk Management, 15(1), p. e12770. doi: https://doi.org/10.1111/jfr3.12770.
Theres, B. L. and Selvakumar, R. (2022) ‘Comparison of landuse/landcover classifier for monitoring urban dynamics using spatially enhanced landsat dataset’, Environmental Earth Sciences, 81(5), pp. 1–8. doi: https://doi.org/10.1007/s12665-022-10242-x.
Townshend, J. R. et al. (2012) ‘Global characterization and monitoring of forest cover using Landsat data: opportunities and challenges’, International Journal of Digital Earth, 5(5), pp. 373–397. doi: https://doi.org/10.1080/17538947.2012.713190.
Uniyal, S. et al. (2022) ‘Quantification of carbon sequestration by urban forest using Landsat 8 OLI and machine learning algorithms in Jodhpur, India’, Urban Forestry & Urban Greening, 67, p. 127445. doi: https://doi.org/10.1016/j.ufug.2021.127445.
Urban, M. et al. (2018) ‘Surface moisture and vegetation cover analysis for drought monitoring in the southern Kruger National Park using Sentinel-1, Sentinel-2, and Landsat-8’, Remote Sensing, 10(9), p. 1482.
Wang, J. et al. (2019) ‘Estimating leaf area index and aboveground biomass of grazing pastures using Sentinel-1, Sentinel-2 and Landsat images’, ISPRS Journal of Photogrammetry and Remote Sensing, 154, pp. 189–201.
Wang, J. et al. (2020) ‘Machine learning-based detection of soil salinity in an arid desert region, Northwest China: A comparison between Landsat-8 OLI and Sentinel-2 MSI’, Science of the Total Environment, 707, p. 136092.
Wang, Q. et al. (2020) ‘Comparative analysis of Landsat-8, Sentinel-2, and GF-1 data for retrieving soil moisture over wheat farmlands’, Remote Sensing, 12(17), p. 2708.
Wang, R., Murayama, Y. and Morimoto, T. (2021) ‘Scenario simulation studies of urban development using remote sensing and GIS’, Remote Sensing Applications: Society and Environment, 22, p. 100474.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
17 Aug, 2024
Editor assigned by journal
08 Aug, 2024
Submission checks completed at journal
08 Aug, 2024
First submitted to journal
02 Aug, 2024

You are reading this latest preprint version

Mapping Green space, Roads & Built-up Areas with Optical Remote Sensing and Polarimetric SAR (A Novel Approach)

Status:

Version 1

Abstract

Figures

Introduction

Materials and Method

Research Area

Methodology

Data collection and creation

Data processing

Results

Discussion and conclusion

Declarations