Major Land Use Transitions and Urbanization Trends

Global trends of accelerating anthropogenic emissions largely contributed by land use and land cover change (LULCC) calls for frequent land use monitoring. The accelerating growth of population drives urban expansion at the expense of forests and agriculture, seriously diminishing the established carbon sinks and food sources. Urban expansion trends seen in the Southeastern U.S. show consistent patterns on the global scale, with continuous unsustainable growth and increasing sprawl that indicate poor land use efficiency. To better monitor land use across time, GIS is used as a tool to support analysis. Using the advantages of GIS, LULCC detection becomes easier as GIS supports analysis on a larger scale and enables users to produce more frequent and up-to-date results. Multiple classification methods, such as supervised classification and neural networks, and indices, such as NDVI, established from an amalgam of LULCC research have helped improve the classification accuracy through time.

Acknowledging the significance of LULCC detection in the field of environmental research, the research objective seeks to provide spatial and temporal patterns of LULCC. The two main research questions of the study are – what are the major land use transitions and what are the urban expansion and sprawl patterns across the Atlanta MSA from 2013 to 2022?

The following table shows the list of data used in this study and also some summarized information on the data.

The study area is set at the Atlanta Metropolitan Statistical Area (MSA), Georgia. As one of the main commercial, industrial, and transportation centers of the Southeastern U.S., Metro Atlanta has expanded greatly since 1990 (Yang & Lo, 2002), and the trends of accelerating population increase and urbanization are worth discovering. To research on the more recent urban trends in the Atlanta MSA, the study time period is set from 2013-2022 to see decadal shifts in land use, urban expansion, and urban sprawl.

Clarification on the extent of Atlanta MSA is crucial as there are different interpretations of what is considered within the MSA. In this study, the Atlanta MSA is considered as the 29-county extent; the 29 counties within Atlanta MSA are Barrow, Bartow, Butts, Carroll, Cherokee, Clayton, Cobb, Coweta, Dawson, DeKalb, Douglas, Fayette, Forsyth, Fulton, Gwinnett, Haralson, Heard, Henry, Jasper, Lamar, Meriwether, Morgan, Newton, Paulding, Pickens, Pike, Rockdale, Spalding, and Walton County.

The following figure is a full flow chart of the complete analysis performed in the study. The top section represents the data pre-processing step conducted on the datasets to ensure they are ready and clean to be used in the analysis. The bottom section following data pre-processing is the data processing step, where the actual research is conducted.

To detect urban expansion, we use two indicators to dissect the trends – Urban Land Cover Percentage (ULC %) and Rate of Urban Expansion (RUE).

Urban Land Cover Percentage (ULC %)

ULC % is essentially the percentage of urban land within the total area, and the equation is:

ULC % = Urban Area (km ² ) / Total Area (km ² )

In the scale of ULC % by county, Urban Land Cover Percentage (ULC %) is the urban area detected in the county divided by the total area of the county. In the scale of the entire Atlanta MSA, the urban area parameter is the total urban area calculated from summing up all the detected urban areas within each county, and the total area parameter is essentially the total area of the MSA. The ULC % indicator can help with the analysis of how much urban area takes up the area of interest and the trends over time provide straightforward information on urban expansion.

Rate of Urban Expansion (RUE)

The equation essentially calculates the rate of change in urban areas over two years:

RUE = [(UA) _i+n  - (UA) _i ] * 100 / [n * (UA) _i ]

where UA = urban area; i+n, 1 = years in comparison; n = interval of the two years

n accounts for when the comparison is between two years that are not consecutive; when the RUE calculation is between two consecutive years, n becomes 1. The unit of RUE is in percentage (%).

Land Consumption Rate (LCR)

In the analysis of urban sprawl, the tables of U.S. Census Bureau annual population estimates are used to calculate the Land Consumption Rate (LCR). The equation of LCR is formulated to calculate the area of urban land per capita (Fenta et al., 2017):

LCR = Urban area (m ² ) / Population

where population = proportion to ULC %

One important note for the population parameter is that it is not simply the total population of the area but rather a proportion of the population equivalent to the proportion of urban area within the area of interest, assuming that the population is equally distributed. Essentially, LCR extracts urban population data from Census population data based on the size of the urban area. Contrary to population density which calculates population per unit of area, LCR looks at the trend of land consumption by calculating the size of the urban area one person resides.

Here presents 2013 & 2022 final classification results of Atlanta MSA from supervised classification followed by NTL verification on urban classification. A total of 4 land types are classified – forest, urban area, water, and non-forest vegetation – and represented by the colors green, red, blue, and yellow, respectively (coded 1-4).

The overall pattern seen across time is that forests remain persistent as the majority of land cover across the entire Atlanta MSA. The urban areas reside at the center of the MSA and spread out with branches to the periphery. Since non-forest vegetation covers both grassland and agriculture, yellow areas that are more persistent over time may suggest agriculture since agricultural land rarely gets transformed back to forest land in such short periods of time (Gibbs et al., 2010), as the conditions for natural regeneration and restoration of forests from agricultural land highly depends on soil conditions and land management.

The following two plots use the classification results from 2013 and 2022 to further present a more visible contrast between the two years. The first plot is a map resulting from subtracting the 2013 classification raster from the 2022 raster to detect pixel-wise changes in land type, specifically focusing on the shifts between forest and non-forest vegetation. The second plot, on the other hand, is a map also resulting from raster subtraction, specifically focusing on the newly added urban areas in 2022 and associating the increase to either pixel of forest or non-forest vegetation in 2013.

The sparse yellow and green patches show a constant exchange of land type between forest and non-forest vegetation at the periphery of the MSA center. Patches that shift between green and yellow may have shifting spectral signatures suggesting that the areas have varying vegetation health and greenness over the years that make them resemble either forests or non-forest vegetation from time to time. Moreover, the raster subtraction between 2013 and 2022 shows that most vegetation-to-urban area transitions happen close to the MSA center, the majority at the expense of forests.

For each of the classification result maps, the quantity of pixels for each land class is extracted and presented in the annual land cover table. The table is a numerical display of the area of each land type throughout the years from 2013-2022.

Comparing among land types, forest is the major land cover of Atlanta MSA, which is also visibly seen from the classification results as green takes up most of the study area, persistently bigger in area than all other three land types combined. Extending from the visual patterns of urban areas suggesting its central location within MSA, the urban land has the second most area. Analyzing the trend in area change for each land type reveals that forests have decreased over time from 2013 to 2022. The net increase in urban and non-forest vegetation throughout the decade may suggest trends of deforestation contributing to these land types’ increase in area, which coincides with literature reviews stating the rapid decline seen in forests in Atlanta (Yang & Lo, 2002). For water that usually remains pretty stagnant in area, the fluctuation in area throughout the decade may suggest possible limitations in classification, since vegetation alongside rivers often shadows the water bodies, and algal level in lakes may cause classification to be misinterpreted as non-forest vegetation.

Following the presentation of classification results and annual land cover data, the section begins with the discussion of the first research objective of the study – detecting major land use transitions within Atlanta MSA from 2013 to 2022.

The transition matrix and the circular flow plot together provide numerical and visual information on land use from 2013 to 2022 in Atlanta MSA. The high numbers on the diagonal of the transition matrix compared to other cells suggest that the majority of the pixels of each land type remain unchanged. From the annual land cover data, forests show a net decrease, and urban areas and non-forest vegetation show net increases in land cover between 2013 and 2022. Taking the trends seen in annual land cover and using them as a basis for further analysis of land use transition provides more complex explanations for the land use patterns seen in Metro Atlanta. Regarding forests, ~11.9% of 2013 forest land is transitioned to other 3 land types, while the majority of the forests (~88.81%) remain as forests.

Regarding annual land cover showing a net decrease in forest area, the transition matrix shows that, excluding forests that remain unchanged, forests contribute the most to urban areas and non-forest vegetation. These extracted patterns from the transition matrix can also be seen in the circular flow plot. The thickness of the green arrow pointing back to the forest section shows that the majority of the forests remain constant. Among the other arrows, the forest arrow that points towards non-forest vegetation is the thickest, therefore restating the pattern seen in the transition matrix that forests transition to non-forest vegetation the most out of the land types other than forest. Past literature reviews have also supported the patterns, discovering that the decline in forest land is largely caused by the rise of agriculture, grassland, or urban areas (Yang & Lo, 2002). What is significant about this finding is understanding where forest land has gone and this enables researchers to quantify and track the negative impacts on climate and ecosystem associated with deforestation.

Regarding the annual land cover showing a net increase in urban area, the transition matrix shows that forests and non-forest vegetation are the two main contributors to the increase, excluding the urban areas that remain unchanged. Transitioned forests make up ~9.49% and non-forest vegetation makes up ~4.68% of the total 2022 urban land area. Looking at the circular flow chart, the extracted information from the transition matrix is shown by the green and yellow arrows pointing towards the urban area. The forest arrow that points toward the urban area is thicker than the yellow arrow, indicating that the transitioned forests make up a higher proportion of the 2022 urban area than non-forest vegetation does. The study’s finding of the sources of urban area increase coincides with literature reviews on the major sources of urban expansion. Within the past research articles, trends seen in the Southeastern U.S. and past Atlanta both show that forests and agricultural lands have rapidly declined in recent decades at the expense of accelerated increase in population, urbanization, and human development (Zhao et al., 2013). The study’s annual land cover and transition matrix providing numerical values and quantity of each land type’s area that underwent a transition to urban land throughout the decade can largely support better future sustainable urban growth.

Regarding annual land cover data showing a net increase in non-forest vegetation, the majority of the transition happens in forests, excluding non-forest vegetation that remained unchanged. The transition matrix shows that forest contributes ~25% of 2022 non-forest vegetation. Looking at the circular flow chart, the thickness of the arrow originating from forest heading towards non-forest vegetation increases as it transitions, indicating that the land transitioned from forest is just a minor proportion but that transitioned land makes up a large proportion of the 2022 non-forest vegetation. Furthermore, two arrows are pointing towards each other between forests and non-forest vegetation, suggesting that there are shifts between the two land types. This finding also can be seen in the classification result maps, and possible explanations include the fluctuation in the spectral signature of certain trees that shift in greenness or vegetation health that makes classifying the land under either forest or non-forest vegetation a difficult task. The transition matrix and circular flow plot together answer the first research question on the major land use trends in Atlanta MSA.

The following section will discuss several discoveries during urban analysis and present answers and explanations for the second research question on urban trends of expansion and sprawl. The discussion focuses on using the indicators of urban expansion and sprawl – ULC %, RUE, and LCR – to dissect the general trends of the entire Atlanta MSA and also inter-county differences within the general trends.

Urban Expansion

The plot is a stacked plot of Urban Land Cover Percentage (ULC %) on the top and the Rate of Urban Expansion (RUE) plot at the bottom; the RUE plot presents the RUE between every consecutive year and also across the decade.

The first takeaway message for the ULC % plot is that through annual fluctuations there is an overall increase of ULC % from 2013 to 2022. The ULC % increased from 25.28% to 26.62% between 2013 and 2022, a 1.34 percentage point increase. ULC % is largely accumulated from 2015 to 2017, indicating a continuous accelerated growth in urban areas during the period of time. For RUE, Atlanta MSA has a RUE value of 0.59, indicating a slow and gradual increase in urban areas across time. Taking into account the RUE plot, the greatest ULC % growth happens between 2021 and 2022 with a 1.8 percentage point increase (24.8% to 26.6%), therefore also having the highest RUE value of 7.27%. On the other hand, the greatest ULC % decline happens between 2020 and 2021 with a 1.9 percentage point decrease (26.7% to 24.8%), therefore also having the lowest RUE value of -7.17%. Possible explanations behind the large fluctuation in urban areas between 2020 and 2022 may allude to the COVID-19 pandemic as city and household light usage has drastically changed and the average monthly NTL is seen to decrease (Liu et al., 2020). Since the classification results also rely on NTL images for accurate urban classification, the fluctuations in average monthly NTL during the pandemic may result in fluctuations in the amount of urban area being classified. The decrease in average monthly NTL will result in below-than-usual classified urban areas, and as the pandemic stretches to 2022 with the deceleration of COVID cases, the urban area classified bounces back with the resumption of activities and light usage in the city.

Urban Sprawl

The following figure shows the Land Consumption Rate (LCR) of Atlanta MSA across the decade. LCR is calculated by dividing the total urban area by the total urban population, therefore indicating the urban area one person resides. One side note to be reiterated is that the urban population for the LCR calculation is the proportion of the total Census population accounted for the proportion of urban area among the entire Atlanta MSA. Urban sprawl by definition is the decreasing density in urban areas; LCR is therefore an efficient indicator of urban sprawl because tracking the values through time tells that increasing values of LCR means that the density among the urban area is decreasing and the urban area one person resides in increases.

The main takeaway message for the LCR plot is that there is an overall decreasing trend across Atlanta MSA. From 2013 to 2022, LCR dropped from 4,254.23 m2/capita to 3,754.74 m2/capita, an 11.74% decrease. This may suggest that the rate of increase for the population is growing faster than the urban expansion rate, causing urban density to decrease over time. Atlanta has immense urban growth in the late 1900s, and the plot suggests that Atlanta MSA’s urban growth is slowing down in the recent decade while the population continues to increase.

Following the discovery of the major urban trends of the entire Atlanta MSA, the analysis proceeds to delve into the intercounty comparison of urban trends within the general trends seen in the MSA. The urban analysis again focuses on using ULC % and RUE to detect trends of urban expansion and LCR to detect trends of urban sprawl, but now at a different scale – the 29 counties within Metro Atlanta.

Urban Expansion

The following figure is the cumulative ULC % plot of each county; the light brown indicates the ULC % in 2013 and the dark brown bar plot is the accumulation of ULC % in 2022 since 2013. In counties that have lower ULC % values in 2022 than in 2013, more transparent dark brown bar plots with dark brown outline is shown instead to show both the 2013 ULC % (light brown bar plot) and the decreased ULC % value in 2022.

The general trend of ULC % across Atlanta MSA is a net increase from 2013-2022 with fluctuations in between. Regarding county-wise ULC %, out of all 29 counties, only 9 counties have decreased in ULC % over the decade (Butts, Clayton, Cobb, DeKalb, Douglas, Jasper, Meriwether, Morgan, and Rockdale County). The highest net cumulative ULC% is in DeKalb County, with 89.52% of urban area. DeKalb County is one of the counties that has a decreasing ULC %, meaning that the ULC % in 2013 is even higher than 89.52%, over 90%. The lowest net cumulative ULC % is in Jasper County, with 0.94% of urban area. Jasper County is also one of the counties with a decreasing ULC % in 2022; the low percentage indicates that the county is highly rural. The largest increase in ULC % between 2013 and 2022 happens in Forsyth County, with a 9.39 percentage point increase, and the lowest increase in ULC % happens in Lamar County, with a 0.11 percentage point decrease. On the other hand, the largest decrease in ULC % happens in Douglas County, with a 1.29 percentage point decrease, and the lowest decrease in ULC % happens in Jasper County, with a 0.18 percentage point decrease.

Following the numerical graph of the ULC % changes, the following thematic maps of 2013 and 2022 ULC % give a visual presentation of the distribution of ULC % among the counties and also across time.

The thematic map provides spatial patterns that may otherwise be more difficult to comprehend in graphs. The high values of ULC % are seen to be located at the center of the MSA, in counties such as Fulton, Cobb, Gwinnett, DeKalb, and Clayton County, and lower values of ULC % are seen at the periphery of the MSA. This coincides with the urban sprawl patterns of density that decrease with the distance away from the city center. Within the general trend of increasing ULC % within the entire Atlanta MSA, this increase is seen expanding to one county southeast of the center (Henry) and two counties northeast (Forsyth and Barrow). Trends seen in the ULC % plot, such as DeKalb County having the highest net cumulative ULC % and Forsyth County having the highest increase in ULC % across the decade, can be visualized in the thematic map.

Furthermore, RUE is calculated for an additional analysis of urban expansion within the 29 counties. The following RUE thematic map reveals spatial patterns that are less conspicuous than ULC %. Northern counties of Atlanta MSA generally have the highest RUE % values, coinciding with higher ULC % values seen in northern counties; the counties at the periphery of MSA also generally have higher RUE. This may be due to the ULC % at the center of MSA already being high and saturated, causing urban growth to occur more at the periphery. Incorporating the ULC % trends, the southeastern counties, Butts, Jasper, and Morgan, and southern county Meriwether remain to be low in ULC % and RUE, indicating overall low urban cover and slow urban growth.

Urban Sprawl

Following urban expansion trends detected through ULC % and RUE, LCR helps provide important insights in urban sprawl within the 29 counties. Similar to the ULC % plot, the LCR plot also reflects the cumulative value from 2013 to 2022; the light brown bars reflect 2013 LCR values and the dark brown bars reflect the increased LCR in 2022 from 2013. Again, the counties with lower LCR in 2022 than in 2013 are shown through more transparent dark brown bars with dark brown outline to fully show the extent of decline in LCR from 2013 to 2022.

The general Atlanta MSA trend in urban sprawl is that LCR is decreasing overall from 2013 to 2022. Regarding county-wise LCR patterns, out of all 29 counties, only 2 counties – Heard and Meriwether County, have increased in LCR; this indicates that within the majority of counties in Atlanta MSA, the urban area one person resides in has decreased, indicating an increase in urban density. The majority of the counties are following the general decreasing trajectory of the entire MSA. Coupled with the ULC % and RUE increasing trend across Atlanta MSA, this decrease in LCR across counties suggests that, although there is urban expansion, the fast population growth is filling up the slowly growing urban areas, therefore increasing urban density. Of the 2 counties that have increasing LCR, Heard County has a higher net cumulative LCR than Meriwether County, with a LCR value of 67,876.33 m2/capita; however, Meriwether County has a higher increase in LCR between 2013 and 2022, with a 2,964.44 m2/capita increase. On the other hand, among the rest of the 27 counties that have decreasing LCR over the decade, DeKalb County has the lowest net cumulative LCR, with a value of 922.69 m2/capita. The highest decrease in LCR happens in Morgan County, with a decrease of 5,863.09 m2/capita, and the lowest decrease in LCR happens in DeKalb County, with a decrease of 80.24 m2/capita. Coupled with trends of low ULC % and RUE in Meriwether County, the county having the highest decrease in LCR suggests that although the urban area is expanding very slowly in Meriwether County, the population growth in the county is causing its LCR values to drop. One notable summary from county-wise trends in urban expansion and sprawl is that DeKalb County is a reoccurring contestant in prominent urban trends. DeKalb County has the highest net cumulative ULC %, lowest net cumulative LCR, and lowest decrease in LCR from 2013 to 2022. This may provide important insights on the county, revealing that DeKalb County has a large urban area within the county and the population density within this large urban area is high.

The following map presents the 2013 and 2022 thematic maps of LCR, providing additional visual insights to the cumulative LCR graph. The general trend across Metro Atlanta shown during the analysis of the entire MSA indicates that there is an overall decreasing trend in LCR. Regarding intercounty differences, low LCR values are located at the center of the MSA, while high LCR values are located at the periphery of the MSA. Among the counties on the periphery of Metro Atlanta, southern counties generally have higher LCR. Coupled with trends of lower ULC % and RUE values seen in southern counties, the high LCR values reveal that southern counties have a small percentage of urban areas and low urban growth within the counties and, within these small urban areas, have also low population. The trend of low LCR at the MSA center ties back to the high ULC % trend seen at the MSA center, and the two together reveal the spatial urban pattern of Atlanta MSA – counties at the center of MSA have a larger percentage of urban areas at tend to be denser in population. This finding is a common pattern of urban expansion, as population density tends to be lower at the periphery of MSA; the urban trends discovered in the study agreeing with common urban patterns also give credibility to the results presented.

Accuracy assessment is performed on the classification maps to show the accuracy of results. The reference data for the assessment is land cover data from the USGS National Land Cover Database (NLCD). Years 2013, 2016, 2019, and 2021 are chosen as the years undergoing accuracy assessment due to data availability from NLCD. The sample size for accuracy assessment is 4000, and 1000 sample points for each land class are sampled through equalized stratified random sampling.

Overall, classification performance is the best for the year 2016, with an overall accuracy of 76.33 and a kappa value of 0.68. Over the four years, forests generally have the highest producer’s accuracy and water has the highest user’s accuracy. Urban area classification generally has consistent producer’s and user’s accuracy of over 70%. In all years, classification results that show non-forest vegetation have a high quantity of pixels being classified as forests in reference data. The mix-up is reasonable due to fluctuating vegetation greenness and health over the year which makes classification that reflects land use across the entire year a challenge. Explanations of this misclassification may either be errors in the classification maps or false-positives in the NLCD dataset. Since NLCD is also an annual land cover dataset, it cannot reflect certain transitions to grassland or changes in greenness across the year. For urban area classification, performance and accuracy are substantial, with spaces for possible improvements, and its lower user’s accuracy compared to non-forest vegetation calls for discussion on the limitations in supervised classification that can explain the performance.

From the decadal analysis, the research discovers overall trends of increasing Urban Land Cover Percentage (ULC %), positive values of Rate of Urban Expansion (RUE), decreasing Land Consumption Rate (LCR), and increasing urban population density across the entire MSA. Within the entire MSA, there are intercounty differences that should be noted. DeKalb County has the highest ULC % and has the highest urban population density (low LCR). Spatially, the periphery of the MSA has low ULC % coupled with high LCR, indicating trends of urban sprawl occurring more often as distance away from the center of the MSA increases. The sources of land classes that contribute to urban growth and the general urban trends seen in Atlanta MSA both are consistent with the patterns seen in the Southeastern U.S. and the global scale, further verifying the results provided in the research.

The land use and urban trends discovered during the research provide a significant basis for further research extending to future projections, environmental justice, and land use management. Climate change remains one of the main challenges of the global stage, and the land cover data can provide good estimates of carbon flux. The urban growth and sprawl trends also give good insights in the current land use efficiency in Atlanta MSA and may act as checkpoints of the urban planning strategies. Although slowing down, urban expansion is still continuing to grow and patterns of urban sprawl are still evident at the peripheries of the MSA. The results call for better sustainable urban growth and more efficient land use policies that can sustain the impacts of climate change.

• Das, S., & Angadi, D. P. (2022). Land use land cover change detection and monitoring of urban growth using remote sensing and GIS techniques: a micro-level study. GeoJournal, 87(3), 2101–2123. https://doi.org/10.1007/s10708-020-10359-1
• Earth Observation Group. (2013). See the World at Night. Payne Institute for Public Policy. https://eogdata.mines.edu/products/vnl/#monthly
• Fenta, A. A., Yasuda, H., Haregeweyn, N., Belay, A. S., Hadush, Z., Gebremedhin, M. A., & Mekonnen, G. (2017). The dynamics of urban expansion and land use/land cover changes using remote sensing and spatial metrics: the case of Mekelle City of northern Ethiopia. International Journal of Remote Sensing, 38(14), 4107–4129. https://doi.org/10.1080/01431161.2017.1317936
• Gibbs, H. K., Ruesch, A. S., Achard, F., Clayton, M. K., Holmgren, P., Ramankutty, N., & Foley, J. A. (2010). Tropical forests were the primary sources of new agricultural land in the 1980s and 1990s. Proceedings of the National Academy of Sciences, 107(38), 16732– 16737. https://doi.org/10.1073/pnas.0910275107
• Güneralp, B., Reba, M., Hales, B. U., Wentz, E. A., & Seto, K. C. (2020). Trends in urban land expansion, density, and land transitions from 1970 to 2010: a global synthesis. Environmental Research Letters, 15(4), 044015. https://doi.org/10.1088/1748- 9326/ab6669
• Hepinstall-Cymerman, J., Coe, S., & Hutyra, L. R. (2013). Urban growth patterns and growth management boundaries in the Central Puget Sound, Washington, 1986–2007. Urban Ecosystems, 16(1), 109–129. https://doi.org/10.1007/s11252-011-0206-3
• Houghton, R. A., & Nassikas, A. A. (2017). Global and regional fluxes of carbon from land use and land cover change 1850–2015. Global Biogeochemical Cycles, 31(3), 456–472. https://doi.org/10.1002/2016GB005546
• Liu, Q., Sha, D., Liu, W., Houser, P., Zhang, L., Hou, R., Lan, H., Flynn, C., Lu, M., Hu, T., & Yang, C. (2020). Spatiotemporal Patterns of COVID-19 Impact on Human Activities and Environment in Mainland China Using Nighttime Light and Air Quality Data. Remote Sensing, 12(10), 1576. https://doi.org/10.3390/rs1210157
• Liu, X., Huang, Y., Xu, X., Li, X., Li, X., Ciais, P., Lin, P., Gong, K., Ziegler, A. D., Chen, A., Gong, P., Chen, J., Hu, G., Chen, Y., Wang, S., Wu, Q., Huang, K., Estes, L., & Zeng, Z. (2020). High-spatiotemporal-resolution mapping of global urban change from 1985 to 2015. Nature Sustainability, 3(7), 564–570. https://doi.org/10.1038/s41893-020-0521-x
• Lwasa, S., Seto, K. C., Bai, X., Blanco, H., Gurney, K. R., Kılkış, S., Lucon, O., Murakami, J., Pan, J., Sharifi, A., & Yamagata, Y. (2022). Urban Systems and Other Settlements. In IPCC Climate Change 2022 - Mitigation of Climate Change. Contribution of Working Group III to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (pp. 861–952). Cambridge University Press. https://doi.org/10.1017/9781009157926.010
• Pesaresi, M., Melchiorri, M., Siragusa, A., & Kemper, T. (2016). Atlas of the Human Planet 2016 - Mapping Human Presence on Earth with the Global Human Settlement Layer.
• Sutton, P. C. (2003). A scale-adjusted measure of “Urban sprawl” using nighttime satellite imagery. Remote Sensing of Environment, 86(3), 353–369. https://doi.org/10.1016/S0034-4257(03)00078-6
• U.S. Census Bureau. (2013). Data - Advanced Search. U.S. Census Bureau. https://data.census.gov/advanced
• U.S. Census Bureau. (2018). Cartographic Boundary Files - Shapefile. U.S. Census Bureau. https://www.census.gov/geographies/mapping-files/time-series/geo/carto-boundary-file.html
• USGS. (2013a). Data (National Land Cover Database). Multi-Resolution Land Characteristics (MRLC) Consortium. https://www.mrlc.gov/data
• USGS. (2013b). EarthExplorer. United States Geological Survey (USGS). https://earthexplorer.usgs.gov/
• Yang, X., & Lo, C. P. (2002). Using a time series of satellite imagery to detect land use and land cover changes in the Atlanta, Georgia metropolitan area. International Journal of Remote Sensing, 23(9), 1775–1798. https://doi.org/10.1080/01431160110075802
• Yasin, M. Y., Abdullah, J., Noor, N. M., Yusoff, M. M., & Noor, N. M. (2022). Landsat observation of urban growth and land use change using NDVI and NDBI analysis. IOP Conference Series: Earth and Environmental Science, 1067(1), 012037. https://doi.org/10.1088/1755-1315/1067/1/012037
• Yasin, M. Y., Mohd Yusoff, M., Abdullah, J., Mohd Noor, N., & Mohd Noor, N. (2021). Urban sprawl literature review: Definition and driving force. Malaysian Journal of Society and Space, 17(2). https://doi.org/10.17576/geo-2021-1702-10
• Zhang, P., Kohli, D., Sun, Q., Zhang, Y., Liu, S., & Sun, D. (2020). Remote sensing modeling of urban density dynamics across 36 major cities in China: Fresh insights from hierarchical urbanized space. Landscape and Urban Planning, 203, 103896. https://doi.org/10.1016/j.landurbplan.2020.103896
• Zhao, S., Liu, S., Sohl, T., Young, C., & Werner, J. (2013). Land use and carbon dynamics in the southeastern United States from 1992 to 2050. Environmental Research Letters, 8(4), 044022. https://doi.org/10.1088/1748-9326/8/4/044022