Previous Article in Journal
New Challenges for Tropical Cyclone Track and Intensity Forecasting in an Unfavorable External Environment in the Western North Pacific—Part II: Intensifications near and North of 20° N
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Atmosphere
Volume 16
Issue 7
10.3390/atmos16070880
atmosphere-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Investigating Urban Heat Islands in Miami, Florida, Utilizing Planet and Landsat Satellite Data

by
Suraj K C
*,
Anuj Chiluwal
,
Lalit Pun Magar
and
Kabita Paudel
College of Agriculture, Health, and Natural Resources, Kentucky State University, 400 E. Main Street, Frankfort, KY 40601, USA
*
Author to whom correspondence should be addressed.
Atmosphere 2025, 16(7), 880; https://doi.org/10.3390/atmos16070880
Submission received: 12 May 2025 / Revised: 4 July 2025 / Accepted: 16 July 2025 / Published: 18 July 2025
(This article belongs to the Section Atmospheric Techniques, Instruments, and Modeling)
Browse Figures
Review Reports Versions Notes

Abstract

Miami, Florida, renowned for its cultural richness and coastal beauty, also faces the concerning challenges created by urban heat islands (UHIs). As one of the hottest cities of the United States, Miami is facing escalating temperatures and threatening heat-related vulnerabilities due to urbanization and climate change. Our study addresses the critical issue of mapping and investigating UHIs in complex urban settings. This study leveraged Planet satellite data and Landsat data to conceptualize and develop appropriate mitigation strategies for UHIs in Miami. Utilizing the Planet satellite imagery and Landsat data, we conducted a combined study of land cover and land surface temperature variations within the city. This approach fuses remotely sensed data to identify the UHI hotspots. This study aims for dynamic approaches for UHI mitigation. This includes studying the status of green spaces present in the city, possible expansion of urban green spaces, the propagation of cool roof initiatives, and exploring the recent climatic trend of the city. The research revealed that built-up areas consistently showed higher land surface temperatures while zones with dense vegetation have lower surface temperatures, supporting the role of urban green spaces in surface temperature reduction. This research can also set a robust model for addressing UHIs in other cities facing rapid urbanization and experiencing mounting temperatures each passing year by helping in assessing LST, land cover, and related spectral indices as well.

1. Introduction

Population growth, industrialization, and urbanization are some of the major anthropogenic factors contributing to rising temperature in urban localities [1]. By 2018, about 55% of the world’s population lived in urban regions, with this figure projected to increase up to 68% by 2050 [2]. While concentration of development, people, and intensive human interactions/activities in specific places enhances social and economic prosperity, it also negatively affects the natural environment and ecosystems [1]. It can be observed that from the late 19th century, with accelerated human interventions, climate change has been increasing on both local and global scales with the increase in average surface temperatures. Tropical cities with the highest urbanization levels concentrate 80–90% of regional populations and frequently exhibit rapid territorial expansion and high deforestation rates [3]. This process significantly affects the ecological balance, environmental health, population well-being and quality of natural resources. Urban areas are expected to expand from 0.6 to 1–3 million km2 between 2015 and 2050, an increase of 78 to 171% over the urban footprint in 2015 [3].
Despite being close to water bodies, Miami, Florida, has its subtropical climate primarily characterized by hot and humid summers followed by mild winters. These kinds of land areas are especially prone to extreme heat and urban heat island effects (UHI) [4]. The study area experiences summer temperatures exceeding 32 °C, worsened by high humidity and sea surface temperature rises contributing to compounded thermal stress [5,6].
As more areas become deforested and urbanized, cities will face intense environmental risks such as dryness, heatwaves, and health risks associated with the unfavorable increased temperatures. To build resilient and sustainable communities, urban planning should incorporate adaptation strategies for tackling these changing conditions. Evaluating the land surface temperature along with the land cover of the urban areas is crucial for understanding the trajectory of environmental changes and to mitigate possible environmental hazards [1,3,4]. The global urban population, which represented 54.5% in 2016, is expected to increase two-thirds by midcentury, indicating increased temperatures from both global warming and urban heat island effects (UHIs in urban areas) [4].
It is critical to evaluate the land surface temperature of the research area to comprehend, investigate, and address issues such as the effects of urban heat islands [5]. The values of the land surface temperature can be used to create an urban heat island raster [4,6]. To fully evaluate the effects of urban heat islands, it is also important to comprehend how land surface temperature, the urban heat island raster, land cover maps, and suitable satellite data are computed. Land surface temperature (LST) is an important variable for understanding the thermal properties of Earth’s surface [7]. LST is the temperature of the Earth’s surface measured by remote sensing devices, such as Landsat 8 thermal bands [8]. The LST raster shows temperature values over the research area, highlighting areas with higher and lower temperatures without considering larger temperature trends or anomalies [9]. The urban heat island (UHI) effect is a well-researched phenomenon where urban localities experience higher temperatures than their rural surroundings. This is because of heightened human activities and urbanization [1,9]. The urban heat island (UHI) raster represents the urban heat island (UHI) effect’s intensity across the study area. Each pixel in the UHI raster contains a value. UHI intensity values indicate temperature deviations from the mean of the entire study area, normalized by its standard deviation [10]. Positive UHI values signify higher temperatures than the mean, suggesting UHI effects, while negative values indicate lower temperatures [11].
Urban ecosystems are diverse and dynamic, with land cover changing quickly because of development and human activity. Monitoring these changes and their consequences, such as urban heat islands (UHIs), is essential for sustainable urban development [12]. Historically, satellite imagery from sources such as MODIS and Landsat has proved useful in this quest. These satellites give significant data on land surface temperature (LST) and land cover, allowing academics and planners to better understand and minimize the negative effects of urbanization. MODIS (Moderate-Resolution Imaging Spectroradiometer) and Landsat satellites have played important roles in environmental monitoring. MODIS, which is carried by NASA’s Terra and Aqua satellites, provides great temporal resolution and captures image data of the Earth’s surface twice a day. This regular monitoring is critical for detecting quick changes in urban environments. MODIS data, with its intermediate geographic resolution, is especially useful for large-scale environmental assessments, such as determining land surface temperature and vegetation health [13]. However, its spatial resolution of 250 m to 1 km may be inadequate for precise, fine-scale measurements [14]. For more than four decades, NASA and the USGS have collaborated on Landsat, which provides high-resolution imagery (30 m) [15]. Landsat’s thermal infrared sensors are essential for measuring LST, making it an important tool for UHI research [16]. Landsat’s long-term data archive is also extremely useful for researching historical land cover changes, long-term patterns in urbanization, and their environmental consequences [17,18,19]. Despite their advantages, MODIS and Landsat have limitations, especially in terms of geographical and temporal resolution. The limitations in spatial and temporal granularity create a gap in detecting fine scales of rapidly changing urban conditions, especially in places like Miami where localized events and their effects might have been missed. This is where Planet data fulfill the need for higher-resolution data. Planet Labs has a vast fleet of small satellites known as Doves that offer daily imagery with a high spatial resolution of 3–5 m [20]. This high frequency and precise detail are especially useful for urban research, where rapid changes in land cover can occur [21]. High-precision remote sensing imagery such as PlanetScope imagery is specifically superior in urban studies, precision agriculture, forestry, disaster management, and climate resilience studies. In UHI studies like these, high-resolution data allows for detailed mapping and fine scales—features that are often missed by coarser sensors. Such fineness is crucial for understanding the variability and implementing local mitigation strategies in the study areas [20,21]. Using Planet data in an area like Miami, Florida, has numerous advantages. Miami is a fast-growing city with considerable environmental issues, such as high temperatures, high humidity, and the possibility of sea-level rises [20,21]. Traditional satellite data, while important, may miss the fine-scale variations and quick changes required for effective urban planning and management [22]. Planet satellite data’s daily digital data can provide up-to-date data on land cover changes, enabling near-real-time monitoring and more responsive decision-making [23,24]. The construction of new buildings, roads, and green spaces can be more carefully monitored [24]. This detailed information is critical for detecting UHI hotspots and developing mitigation solutions, such as the installation of new parks or green roofs [25]. Furthermore, the great temporal precision of Planet data ensures that ephemeral occurrences, such as abrupt land cover shifts, are quickly captured and addressed [25,26].
Collectively, GIS tools and remote sensing data allow us to visualize land surface temperature and urban heat island situations inside a chosen study area. This strategy greatly reduces the urban heat island (UHI) effect by making it possible to identify hotspots, direct the development and construction of green spaces, control vegetation, and put public health measures into action. This study utilizes Planet and Landsat satellite imagery to evaluate land surface temperature, the urban heat island effect, and land cover in Miami. It also studies crucial vegetation indices and classifies areas as extremely sensitive or safe from UHI effects. Using recent Planet and Landsat data from both summer and winter time frames, this study focuses on reducing potential UHI effects in Miami, one of the most densely inhabited and hottest cities in the United States. Several studies have utilized moderate-resolution sensors like Landsat and MODIS to underscore the relationship between vegetation indices and LST/UHI, but high-resolution analysis like this still remains unexplored. This study aims to identify vulnerable temperature areas, evaluate the vegetation cover through spectral indices, and provide evidence-based recommendations for investigating the UHI’s impacts. This study also aims to assess the relationship between vegetation indices and land surface temperatures to understand the spatial and seasonal variations in the urban heat island effects in Miami, Florida. Some of the central research questions of this study include the following: How do vegetation indices vary seasonally in Miami? What are the spatial relationships between vegetation indices and LST in both summer and winter? Which areas of Miami are most susceptible to UHI and how do the patterns differ seasonally?

2. Methods

2.1. Study Area

The study area for our research is Miami, Florida. Miami is the third hottest urban area in the US. The area of Miami is 144 km2. Despite its small size and proximity to water bodies, Miami’s reputation as one of the hottest places in the US suggests the presence of the environmental phenomenon known as the urban heat island (UHI) [27,28]. As one of the smallest land areas among major cities, it hosts a staggering population of 6.18 million. For our study, we just considered the land mass of Miami. Hence, we used shapefiles available from Tiger shapefiles which only considered the land mass. We followed this step to avoid confusion from the impacts of water bodies on the land surface temperature of the study area. The details about the study area can be found below in Figure 1.

2.2. Data Preparation

Land surface temperature was calculated for July and November 2023, from the Landsat 8 Operational Land Imager (OLI) sensor and thermal infrared sensor (TIRS). The Landsat 8 OLI images were downloaded from Earth Explorer. To capture the seasonal variations and vegetation growth patterns, winter and summer images of Landsat were acquired from November and July of 2023, respectively. Only those images which had a lower cloud percentage (<20%) were downloaded for the analysis [29,30]. Landsat L2C2 (Level2 Landsat Collection 2) imagery was used as these images had undergone atmospheric correction [31], resulting in more uniform spectral and geometric properties. These images were later used to compute land surface temperature and urban heat island effects for the summer and winter months of 2023, as stated above. Getis * Ord hotspots for LST were later created for July and November. We aimed to conduct an important investigation of the recent land surface temperature. Regarding the availability of better cloud cover and for the sake of seasonal comparison, we acquired 2 scenes each from July and November for both Landsat and PlanetScope data.
For July, six PlanetScope surface reflectance (8-band) scenes were acquired. The scenes were harmonized using PlanetScope’s Harmonization Tool. This tool is manufactured by Planet Labs and the company headquarter is in San Francisco, California, United States. For November 2023, six surface reflectance (8-band) scenes that were available were downloaded. Like the July data, the scenes were harmonized using the same tool to create a single composite image. All the images were in GeoTIFF format and were clipped and processed using PlanetScope’s capabilities. They were generated as Harmonized PlanetScope-Sentinel-2 (HPS2) products, wherein PlanetScope imagery is resampled to harmonize the Sentinel-2 spectral response. The details of the data can be found in Table 1.
For Landsat data the two best available Landsat scenes from both July and November were acquired for this study. The details of the data can be found in Table 1.
The details of data acquired for this research are presented below in Table 1.

2.3. Spectral Indices and Land Cover Map from PlanetScope Data

The Normalized Difference Vegetation Index (NDVI), Normalized Difference Red Edge (NDRE), and Visible Atmospherically Resistant Index (VARI) were generated for the same months of 2023 from PlanetScope data utilizing Arc GIS pro-3.2.
To make the data analysis and comparison consistent, the nearest available dates to July and November 2023 were used for downloading PlanetScope data. Vegetation indices which were generated from PlanetScope data using ArcGIS pro-3.2 are listed below in Table 2 with their formulas.
Remote sensing indicators affecting surface temperature include the NDVI, NDRE, and VARI. These indicators are used for monitoring the vegetation cover, soil moisture, and surface reflectance, which are directly associated with land surface temperature [32,33,34]. Existing studies show a strong inverse relationship between vegetation indices and LST in most urban areas, where sparse vegetation is linked with higher surface temperatures. Higher values of the Normalized Difference plant Index (NDVI) indicate denser and healthier plant cover. The NDVI is used to evaluate the presence and health of vegetation [34]. Lower Normalized Difference Red Edge Index (NDRE) values, on the other hand, can be a sign of stressed or sparse vegetation, which could exacerbate the impacts of urban heat islands (UHIs) by decreasing transpiration and providing shade [35]. As a result of limiting evaporative cooling and increasing heat absorption, reduced Visible Atmospherically Resistant Index (VARI) values may also indicate impermeable surfaces and sparse vegetation, which would worsen the consequences of UHIs [35].
Utilizing the Planet data for summer, i.e., June, which had better clarity, land cover maps of the study area were also generated. For land cover map construction, the random tree classifier in ArcGIS pro-3.2 was used. The study was classified into two classes: Developed and Green. For each class, 250 training samples were created using the training sample manager tools of ArcGIS pro-3.2. After classification was performed, to assess the accuracy, reference data points were taken from the same map. For each class about 50 random reference points were taken. User and producer accuracies and Kappa statistics were computed for the classified map. Although land cover classification was not the objective of this study, it was generated to visually inspect the land cover scenario of the study area. The study area was classified as Developed and Green using the RF classifier in ArcGIS pro-3.2.

2.4. Land Surface Temperature and Urban Heat Island Raster Using Landsat 8 OLI Data

Land surface temperature was computed utilizing the Landsat images for June and November of 2023 separately. Land surface temperature from Landsat 8 OLI was calculated utilizing the following equations [36]:
Land Surface Temperature (LST) = Tb/[1 + (α ∗ Tb/C2) ∗ ln(ε)]
where Tb in °C (3) is the satellite brightness temperature, α is the wavelength of emitted radiance, C2 = 1.4388 is a constant, and ε denotes emissivity as described in Suribabu et al. [37].
Brightness temperature is calculated according to the following [38]:
Tb = (K2/(ln (K1/L) + 1)) − 273.15
and [37]
ε = 0.004 ∗ Pv+ CV
where K1 is sensor-dependent calibration constant 1 (774.8853) and K1 is sensor-dependent calibration constant 2 (1321.0789), L is Top of Atmosphere (TOA) spectral radiance, Pv is the proportion of vegetation, and CV is the correction value for Landsat images (0.986) [39].
Pv = ((NDVI − NDVImin)/(NDVImax − NDVImin))2
All these operations were performed utilizing the raster calculator tool in ArcGIS pro-3.2.
The graphical workflow used for calculating land surface temperature from Landsat 8 is provided below in Figure 2.
Urban heat islands are areas with above-average LSTs. Urban heat island intensity is defined as the metric for calculating anomalies of temperature within urban areas [11].
UHI   Intensity   standardized = L S T − M e a n ( L S T ) S D ( L S T )
Urban heat island intensity was created for July and November for the study area.

2.5. Relationship Analysis and Suitability Analysis for UHI-Prone Areas

Spectral indices, i.e., VARI, NDVI, and NDRE for winter and summer, as mentioned above, were reclassified. The LST rasters for winter and summer were also reclassified. A total of 100 points were generated on the spectral indices and LST and the values were extracted. At the same points/places, spectral index values were also extracted.
Correlation matrices were created for spectral indices versus LST values for winter, and the same procedure was conducted for summer. Regression analysis was also performed similarly for winter and summer. Statistical analyses were conducted in Microsoft Excel- M365 Excel 16.77–16.80 and Python- 3.12.
For detecting the UHI-prone and less susceptible areas, suitability analysis using the weighted overlay method was performed. LST was given 40%, and the NDVI, NDRE, and VARI indices were each given 20% control in the model. Suitability analysis was performed for both summer and winter periods. This weighted overlay method delineates the UHI-prone areas, reflecting the influence of each variable in contributing to thermal anomalies. Land surface temperature is the primary representative and indicator of surface heating and urban heat intensity. Hence, it was assigned 40%. The indices were provided with 20% each, recognizing their secondary roles in urban heat moderation through evapotranspiration and surface albedo. Similar approaches were found to be practiced in other studies [11,18,34,35].
The detailed study workflow is displayed below in Figure 3.

3. Results

3.1. Spectral Indices of Study Area

The spectral indices for Miami were computed for the winter (November) and summer (July) of 2023. Figure 4, Figure 5 and Figure 6 display the VARI (Visible Atmospherically Resistant Index), NDRE (Normalized Difference Red Edge Index), and NDVI (Normalized Difference Vegetation Index) for both seasons. Figure 4 illustrates the VARI values during winter, ranging from −0.4 to 0.59, and summer, ranging from −0.71 to 0.37. Despite some areas in Miami showing increased greenness in summer, overall greenness is more prominent in the winter VARI images.
Figure 5 depicts the NDRE values for the study area in winter (−0.99 to 0.99) and summer (−0.14 to 0.62). The summer NDRE maps highlight areas with dense and healthy vegetation that were not as apparent in the VARI images. Figure 6 shows the NDVI maps, representing winter (−0.95 to 0.99) and summer (−0.1 to 0.76) scenarios. The NDVI maps indicate that greenness is more pronounced during the summer months. Overall, the spectral index maps for the NDVI, NDRE, and VARI reveal that green vegetation is dominant during the summer period. Some areas with sparse vegetation in summer show noticeable growth in vegetation in winter.

3.2. Land Surface Temperature and Urban Heat Island Raster of Study Area

Figure 7 displays the land surface temperature (LST) for both winter and summer in the study area, while Figure 8 presents the urban heat island (UHI) raster for the same periods in 2023. In the winter LST raster, temperatures ranged from 12.95 to 33.92 °C, contrasting with the summer LST raster where values ranged from 36.27 to 40.83 °C. Higher LST values are notably more prevalent in summer compared to winter, even in areas near water bodies where differences between summer and winter LSTs are discernible.
Figure 8 illustrates the UHI rasters for summer and winter in 2023. During summer, UHI values ranged from 12.83 to 18.99 and from −1.97 to 12.82, indicating both higher and lower values across the study area. In contrast, the UHI raster for winter showed lower value ranges from −32.61 to −30.82 and from −35.37 to −32.62. Specifically, in the southwestern and eastern parts of Miami during winter, lower UHI values were observed compared to summer.

3.3. Land Surface Temperature Gi_Bin Hotspots

Figure 9 depicts the land surface temperature (Gi_Bin) hotspots during the winter and summer of 2023 in Miami, FL. During the summer, hotspots are notably concentrated in the northern and western parts of Miami, whereas they appeared less frequently in the winter. Cold spots were observed predominantly near water bodies and at the periphery of the study area. In summer, hotspots were denser and more concentrated in the central part of Miami, whereas in winter, they appeared sparser in the same locations. The number of cold spots increased during winter compared to summer.

3.4. Land Cover Classification and Accuracy Assessment of Classification

Figure 10 illustrates the land cover classification of the study area, categorized into two primary types: Developed and Green. The classification utilized PlanetScope data from July 2023. To generate the training sample basemaps, both July and November 2023 PlanetScope data was employed. For the Developed category, training samples were selected from areas featuring buildings, roads, bridges, houses, and concrete structures. Conversely, the green category samples were sourced from forests, parks, and green spaces. A total of 350 samples were collected for each category.
The classification was carried out using the random tree classifier, a method within the random forest classification techniques available in ArcGIS Pro-3.2. The classifier’s default settings included a maximum of 50 trees, a maximum tree depth of 30, and a maximum of 1000 samples per class. After the classification process, the area of Developed land was calculated to be 71.86 km2, while green spaces covered 26.67 km2. The total classified area amounted to 98.53 square kilometers. Miami’s total area, including water bodies, is 145.23 km2; the land area of Miami was estimated to be approximately 94 km2. The land cover map indicates that the study area encompasses 98.53 km2 of land, which can be observed in Table 3.
For accuracy assessment, a stratified random sampling method was employed. Fifty validation samples (accuracy points) were generated for both the Developed and Green classes on the classified map. Visual inspection was conducted using basemaps, Landsat images from the same year, and Planet data from both the summer and winter of 2023 as references. A confusion matrix was computed for the land cover classification. The resulting User Accuracy and Producer Accuracy for the Developed class were 0.94 and 0.89, respectively, which is shown in Table 4. For the Green class, the User Accuracy and Producer Accuracy were 0.88 and 0.93, respectively, as shown in Table 4. The Kappa coefficient for the classification was computed as 0.82.
The central urban core area corresponds to Developed land cover. Green spaces are represented by areas with lower urbanization. Vegetated areas act as thermal buffers, especially along coastal parks. The proximity to water areas might play a significant factor here, modulating the UHI intensity.

3.5. Correlation Matrices Between Vegetation Indices and Land Surface Temperature

Figure 11 and Figure 12 represent the correlation matrices between vegetation indices and land surface temperature for the summer and winter seasons, respectively. A total of 100 random points were generated on the raster of LST, VARI, NDRE, and NDVI for the summer and winter periods, and values were extracted for the same points for two time periods.
In Figure 11, the correlation matrix displays the relationship between spectral indices and land surface temperature values of 100 points used for the summer season. As expected, vegetation indices showed stronger association between each other in the correlation matrix. The Pearson correlation coefficients (r) for the VARI with the NDRE and NDVI were found to be 0.67 and 0.70, respectively. The correlation between the NDVI and NDRE was found to be 0.99. The correlations between LST for summer and the NDVI, NDRE, and VARI were found to be negatively associated, with values of −0.34, −0.37, and −0.37, respectively, demonstrating an inverse relationship. With increasing temperature, the inverse relationship between vegetation indices and LST for summer was expected.
Figure 12 presents the correlation matrix between spectral indices and land surface temperature values of 100 points used for the winter season. Vegetation indices showed stronger association between each other in the correlation matrix for winter as well. The Pearson correlation coefficients (r) for the VARI with the NDRE and NDVI were found to be 0.64 and 0.67, respectively. The correlation between the NDVI and the NDRE was found to be 0.98. Correlations between LST for the winter season and the NDVI, NDRE, and VARI were found to be negatively associated, with values of −0.37, −0.39, and −0.53, respectively, demonstrating the inverse relationship. The inverse relationship between spectral indices and LST values was stronger in the winter season than in the summer season.

3.6. Linear Regression Between Vegetation Indices and Land Surface Temperature

Scatter plots showing the regression between vegetation indices and land surface temperature (LST) for the winter and summer seasons, respectively, are shown in Figure 13 and Figure 14. Across the raster data of the LST, VARI, NDRE, and NDVI for both the summer and winter seasons, a total of 100 randomly chosen points were taken. Following that, values corresponding to these points were extracted and subjected to examination for each season.
Figure 13 displays the regression between LST and the VARI, NDRE, and NDVI, respectively, for the winter season. The regression values R2 for regression between LST and the VARI, NDRE, and NDVI for the 100 points used in this study for winter were found to be 0.28, 0.15, and 0.14, respectively. Figure 14 presents the scatter plots showing the regression between LST and the VARI, NDRE, and NDVI, respectively, for the summer season. The regression values R2 for regression between LST and the VARI, NDRE, and NDVI for the 100 points used in this study for summer were found to be 0.14, 0.13, and 0.12, respectively.

3.7. Suitability Analysis of Urban Heat Island Intensity Effect in Study Area

Figure 15 displays the suitability analysis for urban heat island intensity for the winter and summer seasons for the study area. The raster computed for the suitability analysis shows the UHI-prone areas in five different classes, namely not prone, less prone, mildly prone, prone, and highly prone areas denoted with dark blue, blue, yellow, brown, and red, respectively. The UHI winter raster can be seen in the left part and the UHI summer raster can be seen in the right part of Figure 15.
The dominant presence of red, yellow, and brown color patches was observed in the UHI summer raster’s central portion. Only in the coastal areas and some discontinuous patches was blue observed in some eastern parts of the study area. The results indicate that central portions where a higher population and more mobility, infrastructure, and development are evident are experiencing a greater presence of UHI-prone areas, whereas coastal areas are seen as areas less prone to urban heat island effects. The central portion of the study area shows the dominant presence of highly prone areas, indicating prominent UHI effects. The built-up areas absorb and retain more heat, leading to increased land surface temperatures during summertime. Notably, in winter, these areas appear less widespread and intense compared to summer. The dominant presence of red, yellow, and brown color patches in UHI winter raster’s central portion can be observed, which aligns with the observations of the UHI summer raster. In the UHI winter raster, prominent blue patches can be seen in the coastal areas, especially in the west, north, and east parts of the study area, which were not present in the UHI summer raster. The coastal areas experienced significant UHI effects during summer and cooling in winter. Season variations, atmospheric conditions, and human activities are some of the influential factors that impact the intensity and spatial distribution of UHIs. This affirms that the study area is experiencing pronounced UHI effects during summer in those areas of the study area.

4. Discussion

The results of our study present interesting details about the study area. With the use of Planet and Landsat data, our findings center on spectral indices, land surface temperature, urban heat islands, land surface temperature Gi_Bin hotspots, land cover classification, correlation and regression analysis between vegetation indices and land surface temperature, and suitability analysis for reducing the intensity of the urban heat island effect in the study area. This study utilized Landsat 8 Operational Land Imager (OLI) and TIRS and PlanetScope scene (PSS) data for July and November of 2023, accounting for probable seasonal variations in the study area. Vegetation indices and land cover maps were extracted from PlanetScope data, whereas land surface temperature was extracted using Landsat OLI data.
The analysis of spectral indices, i.e., VARI, NDRE, and NDVI, for the study area in the winter and summer of 2023 explored the seasonal vegetation dynamics of Miami, Florida. These indices present noticeable differences in vegetation health, greenness, and density across the two seasons, which aids in broadening our understanding of environmental and climatic influences on vegetation. The VARI values present an unprecedented trend where greenness was found to be more pronounced in winter than in summer, which contrasts with what is typically expected in temperate climates. The values of the VARI range from −0.4 to 0.59, whereas the values of the VARI in summer range from −0.71 to 0.37. The prominent winter greenness indicated by higher VARI values could be attributed to Miami’s unique subtropical climate, whereas the cooler winter months might be more conducive to facilitating better vegetation health compared to the hot and drought-prone summers [40]. Reduced moisture and spiked summer temperatures are generally considered stressful to vegetation, resulting in lower greenness values, as observed. The NDRE values, on the other hand, present a different scenario. In summer, the NDRE values range from −0.14 to 0.62, which indicates denser and healthier vegetation than in winter where the values range from −0.99 to 0.99. This index is especially sensitive to the chlorophyll content in the red edge spectral region, capturing the vegetation even when it might not be visible in VARI images [41]. The higher NDRE values in summer likely indicate the presence of more robust photosynthetic activity in certain areas of Miami, especially regions with healthy vegetation that can sustain higher temperatures and less moisture. The NDRE, on other hand, explores a different perspective. In summer, NDRE values range from −0.14 to 0.62, which suggests denser and healthier vegetation than in winter where the values range from −0.99 to 0.99. The NDRE, which is sensitive to chlorophyll content in the red edge spectrum, captures vegetation that might not be noticeable in the VARI images. Higher NDRE values in summer indicate the presence of higher photosynthetic activity in certain areas of Miami, especially in regions with vegetation that can withstand higher temperatures and lack of moisture. The NDVI images indicate the pronounced greenness pattern, with summer NDVI values ranging from −0.1 to 0.76 compared to the winter range of −0.95 to 0.99 [42]. While the summer season generally shows higher NDVI values, hinting at more vigorous vegetation growth, we could see areas in winter with considerable green coverage. The difference in observation between the VARI and NDVI in summer months suggests that while the NDVI is capturing border vegetation dynamics, the VARI may be more sensitive to the influence of atmospheric situations and soil background reflectance, which can mask the true nature of greenness in the summer imagery [43].
The land surface temperature (LST) and urban heat island (UHI) raster analyses for Miami in 2023 present distinct seasonal variations with significant differences in the winter and summer. The LST raster in Figure 7 indicates a sharp contrast between winter and summer conditions. In winter, the LST ranged from 12.95 to 33.92 °C, with milder temperatures across the region [5]. The LST at the start of summer, however, showed considerably higher values ranging from 36.27 to 40.83 °C. This increment, especially in urban areas, suggests that heat-absorbing surfaces like asphalt, concrete, and buildings overrun the natural cooling effect of nearby water bodies. The evident increase in summer temperatures is obvious given Miami’s subtropical climate, but the higher LST values even in areas adjacent to water bodies suggest that heat retention in urban regions may overshadow the moderating cooling effect of water during hotter seasons [44,45]. This could mean the presence of heat-absorbing surfaces such as asphalt, concrete, and buildings contributes to higher LST values and worsens the urban heat effect during the summer season [44]. The UHI analysis in Figure 8 further reinforces these findings. During summer, UHI values ranged from 12.83 to 18.99, indicating significant heat island formation for many parts of the study area. The presence of spiked UHI values suggests the urban areas are intensifying the heat stress in Miami, especially during the hottest months [46]. In contrast, UHI values during winter ranged from −32.61 to −30.82 in the southwestern and eastern parts of the city. The role of water bodies is crucial in shaping Miami’s UHI patterns. Areas near the coastlines and water features consistently showed lower LST and UHI values, underscoring the cooling effects of water bodies. These cold spots buffer the temperature and generate cooling effects. The lower UHI values during winter indicate that in cooler months the urban environment has less impact on temperature elevation. This might be due to reduced solar radiation and reduced heat accumulation. The spatial distribution of UHI values also highlights interesting patterns. In summer, some areas faced both higher and lower UHI values (−1.97 to 12.82), presenting a mix of urban heat pockets and cooler zones [5]. The cooler zones could be due to green spaces, parks, and proximity to water bodies that provide some neutralizing effect on the intense heat [44,45,46]. In winter, however, lower UHI values dominate, which shows that the built environment retains less heat during cooler periods while natural aspects like vegetation and water bodies might also contribute to overall temperature moderation.
The land surface temperature (LST) hotspot analysis and land cover classification results in Figure 9 and Figure 10, respectively, reveal important seasonal and spatial patterns in Miami, Florida, that highlight the effects of urbanization on temperature patterns. In summer, LST hotspots can be observed in the northern, western, and central areas of the study areas where urban areas with more developed regions such as roads, buildings, and infrastructures absorb and retain more heat [46]. This observation indicates the intensification of the urban heat island (UHI) effect [47]. In contrast, during winter, these hotspots are fewer and less intense, while cold spots, primarily near water bodies and peripheral regions, are more pronounced, showing the natural cooling effects of water and green spaces in the study area [46]. The land cover classification map in Figure 10 and Table 3 shows that 71.86 km2 (76%) of the study area is found to be developed, while only 26.67 km2 (27%) is classified as green space. The dominance of developed land with sparse green spaces is an obvious major contributor to the UHI effect, particularly during hotter summer months when urban surfaces absorb more heat [5]. The high Kappa coefficient (0.82) and the accuracy metrics for the Developed and Green land cover, as shown in Table 4, verify that the random forest classification employed for classification is reliable, providing higher confidence in the spatial patterns obtained.
The correlation and regression analysis between vegetation indices (VARI, NDRE, NDVI) and land surface temperature (LST) for both the summer and winter seasons provide valuable context on the relationship between vegetation health and land surface temperatures in the study area. Figure 11 and Figure 12 present the correlation matrices for summer and winter, respectively, showing a consistent trend where vegetation indices are positively correlated with each other, with a Pearson correlation coefficient of 0.67 to 0.70 in summer and 0.64 to 0.67 in winter for the VARI with the NDRE and NDVI, which was similar to results found in research performed in a tropical city in India [48]. Notably, the NDVI and NDRE exhibited a near-perfect correlation (r = 0.99 for summer and 0.98 for winter), indicating their strong association with each other across both seasons, which was expected, as in a report from a previous study [49]. In terms of LST, the correlation with vegetation indices was negative, as was expected due to the inverse relationship between land surface temperature and vegetation cover, which aligned with similar research performed by Sharma et al. [50]. In summer, the correlation of LST with the NDVI, NDRE, and VARI ranged from −0.34 to −0.37 while in winter the inverse relationship was slightly stronger, with correlations ranging from −0.37 to −0.53, suggesting an enhanced cooling effect of vegetation during the cooler months [51]. Scatter plots of regression between vegetation indices and land surface temperature (LST) for the winter and summer seasons are presented in Figure 13 and Figure 14, respectively. In the winter season, shown in Figure 13, the regression values (R2) for LST with the VARI, NDRE, and NDVI were found to be 0.28, 0.15, and 0.14, respectively. This makes it evident that the VARI has the strongest explanatory power for LST variability during winter, suggesting greenness might be significant in regulating surface temperatures during colder months [48]. The moderate regression (R2) values for the NDRE and NDVI present that the red edge and overall vegetation health have less influence on temperature variability compared to the visual green captured by the VARI. In contrast, the summer season regression, shown in Figure 14, showed weaker associations between LST and vegetation indices, with R2 values of 0.14, 0.13, and 0.12 for the VARI, NDRE, and NDVI, respectively. These lower R2 values suggest that the influence of vegetation on LST is reduced during summer when temperature might overshadow the cooling effects of vegetation, probably due to enhanced urban heat retention [52]. These seasonal differences in regression suggest that vegetation is more effective in mitigating land surface temperatures during winter periods, as indicated by higher regression values in winter than in summer [53]. This seasonal difference underscores the importance of vegetation for urban temperature regulation, especially in colder months [53,54].
The suitability analysis for urban heat island (UHI) intensity, as shown in Figure 15, presents distinct seasonal patterns in UHI-prone areas for both winter and summer. The raster maps classified into five categories, namely ‘not prone’, ‘less prone’, ‘mildly prone’, ‘prone’, and ‘highly prone’, indicate that the central portion of the study area is dominated by urban infrastructure and dense population, making it highly susceptible to UHI effects during both the seasons [55]. In the summer UHI raster, the prominent presence of red, yellow, and brown patches in the central part can be seen, highlighting areas with enhanced UHI intensity, where development, mobility, infrastructure, and human activity are concentrated [56]. Coastal peripheral areas, on the other hand, mostly showed blue patches, indicating that these regions are less prone to UHI effects, likely due to proximity to water bodies and the cooling effect generated by the water bodies [57]. The winter UHI raster follows a similar pattern, with red and yellow patches dominating the central portion of the study area, highlighting that these areas experience heightened UHI effects throughout the year due to enhanced infrastructure, population, and mobility [58]. However, a notable difference is the appearance of blue patches particularly in the northern, eastern, and western coastal areas, indicating a lower UHI intensity during winter in these regions compared to summer, which might be because of enhanced coastal effects, enhanced impact of vegetation on land surface temperature, and more heat retention during the winter season [59]. A similar study performed on urban heat island intensity in Miami also elucidates similar results to ours. It was observed that average UHI intensity and concentration of UHI days increased across all seasons during the analysis of UHIs in the study [60]. Our findings showed that throughout both the winter and summer months examined in our study, UHI intensity was higher in locations nearer to artificial land cover than in natural features like green spaces and water bodies. A study undertaken in South Florida found similar results [61]. PlanetScope data from November and July, which correspond to the winter and summer seasons, respectively, were used in our investigation. Our research provides a thorough and unique method by combining technologies including geospatial enrichment, hotspot analysis, and suitability mapping. Although we made every effort to utilize the available data and achieve our research objectives within the given time constraints, we recognize certain limitations. Although this study was focused in 2023, there is a difference in date acquisition between the PSS scenes and Landsat 8 OLI. Water data was excluded from this study but can be incorporated in future studies. Use of hyperspectral imagery and drone imageries can further enhance the quality of similar studies in the future. Incorporating multiple years of data could be useful for studying the trend of the urban heat island effect in the study area.

5. Conclusions

This study explored the relationship between spectral indices and land surface temperature for Miami, Florida. The land cover map was generated for the study area with two major classes: Developed and Green. This study focused on conducting suitability analysis for the urban heat island effect. Land surface temperature hotspots were generated. The urban heat island intensity and land surface temperature raster for the summer and winter of 2023 were also generated. This study uses the recently less used Planet Scope scene data and Landsat 8 OLI images for 2023. The land cover classification and spectral indices were generated from Planet Scope data due to the better resolution and quality of the data. The land surface temperature and urban heat island intensity raster were generated using Landsat 8 OLI images. Weighted overlay analysis was used to perform a suitability analysis. In the model, the NDVI, NDRE, and VARI indices received 20% control each, whereas LST received 40%. Suitability analysis was performed for both the summer and winter periods. There have been numerous similar previous studies, but our study uses Planet Scope scene data for generating the spectral indices and for land cover classification. Use of Landsat 8 OLI data for both the winter and summer seasons of 2023 is itself a notable experiment. Planet data has been rarely used in these kinds of studies; hence, this opens a further scope for their usage. The suitability analysis conducted in this study helped us to visualize the different classes in Miami, Florida, based on the susceptibility to urban heat island effect intensity. However, a few limitations of this study should be noted. One was the temporal mismatch between the Planet Scope and Landsat 8 image acquisition dates, which might contribute to some seasonal bias. Another is the classification scheme, which was simplified into only two land cover classes, which might have oversimplified urban heterogeneity.
Despite this, this research contributes to a novel integration of Planet data for high-resolution UHI investigation and presents a replicable geospatial framework for hotspot identification and seasonal analysis. This research is focused on urban heat island effect intensity studies and helping with mitigating those effects. The suitability analysis raster for both winter and summer helps us to visualize how the UHI effect differs seasonally as well. Based on the findings, prioritization of green infrastructure in heat-prone urban centers and protected and enhanced vegetation in densely built-up areas should be encouraged. Municipalities should consider incorporating high-resolution commercial satellite data into their urban climate monitoring programs. Based on the suitability analysis and LST hotspot mapping, urban planning should be focused on prioritizing the afforestation and expansion of green infrastructure in high-risk zones, especially in central Miami where the UHI intensity is elevated all year round. These suggestions are based on geospatial data, and this approach can be integrated into zoning and climate-resilient urban planning projects.

Author Contributions

Conceptualization, S.K.C., A.C. and K.P.; Methodology, S.K.C., L.P.M. and K.P.; Software, S.K.C.; Investigation, S.K.C. and A.C.; Resources, A.C. and K.P.; Writing—review & editing, L.P.M. and K.P.; Visualization, L.P.M. All authors have read and agreed to the published version of the manuscript.

Funding

This research is funded by the USDA-NIFA Evans–Allen Grant # 7004460 to the Kentucky State University.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Acknowledgments

The authors are grateful to Sudip Poudel, Deepak Khatri, and anonymous reviewers for their support during the research and revision of the manuscript.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Aslan, N.; Koc-San, D. The use of land cover indices for rapid surface urban heat island detection from multi-temporal landsat imageries. ISPRS Int. J. Geo Inf. 2021, 10, 416. [Google Scholar] [CrossRef]
  2. United Nations. 2018 Revision of World Urbanization Prospects. 2018. Available online: https://www.un.org/en/desa/2018-revision-world-urbanization-prospects (accessed on 23 June 2023).
  3. Carrillo-Niquete, G.A.; Andrade, J.L.; Valdez-Lazalde, J.R.; Reyes-García, C.; Hernández-Stefanoni, J.L. Characterizing spatial and temporal deforestation and its effects on surface urban heat islands in a tropical city using Landsat time series. Landsc. Urban Plan. 2022, 217, 104280. [Google Scholar] [CrossRef]
  4. Kamali Maskooni, E.; Hashemi, H.; Berndtsson, R.; Daneshkar Arasteh, P.; Kazemi, M. Impact of spatiotemporal land-use and land-cover changes on surface urban heat islands in a semiarid region using Landsat data. Int. J. Digit. Earth 2021, 14, 250–270. [Google Scholar] [CrossRef]
  5. Amindin, A.; Pouyan, S.; Pourghasemi, H.R.; Yousefi, S.; Tiefenbacher, J.P. Spatial and temporal analysis of urban heat island using Landsat satellite images. Environ. Sci. Pollut. Res. 2021, 28, 41439–41450. [Google Scholar] [CrossRef] [PubMed]
  6. Xian, G.Z. Monitoring and Assessing Urban Heat Island Variations and Effects in the United States; US Geological Survey: Reston, VA, USA, 2021. [Google Scholar] [CrossRef]
  7. Mueller, C.; Hussain, R.; Xian, G.Z.; Shi, H.; Arab, S. Mapping the Surface Urban Heat Island effect using the Landsat Surface Temperature Product. In Proceedings of the IGARSS 2023—2023 IEEE International Geoscience and Remote Sensing Symposium, Pasadena, CA, USA, 16–21 July 2023. [Google Scholar] [CrossRef]
  8. Li, Z.L.; Tang, B.H.; Wu, H.; Ren, H.; Yan, G.; Wan, Z.; Trigo, I.F.; Sobrino, J.A. Satellite-derived land surface temperature: Current status and perspectives. Remote Sens. Environ. 2013, 131, 14–37. [Google Scholar] [CrossRef]
  9. Tran, D.X.; Pla, F.; Latorre-Carmona, P.; Myint, S.W.; Caetano, M.; Kieu, H.V. Characterizing the relationship between land use land cover change and land surface temperature. ISPRS J. Photogramm. Remote Sens. 2017, 124, 119–132. [Google Scholar] [CrossRef]
  10. Chakraborty, T.; Hsu, A.; Manya, D.; Sheriff, G. A spatially explicit surface urban heat island database for the United States: Characterization, uncertainties, and possible applications. ISPRS J. Photogramm. Remote Sens. 2020, 168, 74–88. [Google Scholar] [CrossRef]
  11. Kim, S.W.; Brown, R.D. Urban heat island (UHI) intensity and magnitude estimations: A systematic literature review. Sci. Total Environ. 2021, 779, 146389. [Google Scholar] [CrossRef] [PubMed]
  12. Xuan, S.; Ning, Q.; Lei, Z. Impacts of the land use transition on ecosystem services in the Dongting Lake area. Front. Environ. Sci. 2024, 12, 1422989. [Google Scholar] [CrossRef]
  13. Meier, J. Global Irrigation Mapping-the Role of Spatial Resolution of Current and Future Earth Observation Missions. Ph.D. Thesis, Ludwig-Maximilians-Universität München, Munich, Germany, 2023. [Google Scholar]
  14. Mishra, D.R.; Ghosh, S. 14 Using Moderate-Resolution Satellite Sensors for Monitoring the Biophysical Parameters and Phenology of Tidal Marshes. Remote Sens. Wetl. Appl. Adv. 2015, 283. [Google Scholar]
  15. Wulder, M.A.; Roy, D.P.; Radeloff, V.C.; Loveland, T.R.; Anderson, M.C.; Johnson, D.M.; Healey, S.; Zhu, Z.; Scambos, T.A.; Pahlevan, N.; et al. Fifty Years of Landsat Science and Impacts. Remote. Sens. Environ. 2022, 280, 113195. [Google Scholar] [CrossRef]
  16. Jimenez-Munoz, J.C.; Sobrino, J.A.; Skoković, D.; Mattar, C.; Cristobal, J. Land surface temperature retrieval methods from Landsat-8 thermal infrared sensor data. IEEE Geosci. Remote Sens. Lett. 2014, 11, 1840–1843. [Google Scholar] [CrossRef]
  17. Cristóbal, J.; Jiménez-Muñoz, J.C.; Prakash, A.; Mattar, C.; Skoković, D.; Sobrino, J.A. An improved single-channel method to retrieve land surface temperature from the Landsat-8 thermal band. Remote Sens. 2018, 10, 431. [Google Scholar] [CrossRef]
  18. Fu, P.; Weng, Q. A time series analysis of urbanization induced land use and land cover change and its impact on land surface temperature with Landsat imagery. Remote Sens. Environ. 2016, 175, 205–214. [Google Scholar] [CrossRef]
  19. Dewan, A.M.; Yamaguchi, Y. Land use and land cover change in Greater Dhaka, Bangladesh: Using remote sensing to promote sustainable urbanization. Appl. Geogr. 2009, 29, 390–401. [Google Scholar] [CrossRef]
  20. Frazier, A.E.; Hemingway, B.L. A technical review of planet smallsat data: Practical considerations for processing and using planetscope imagery. Remote Sens. 2021, 13, 3930. [Google Scholar] [CrossRef]
  21. Csillik, O.; Kumar, P.; Mascaro, J.; O’Shea, T.; Asner, G.P. Monitoring tropical forest carbon stocks and emissions using Planet satellite data. Sci. Rep. 2019, 9, 17831. [Google Scholar] [CrossRef] [PubMed]
  22. Wu, H.; Gui, Z.; Yang, Z. Geospatial big data for urban planning and urban management. Geo Spat. Inf. Sci. 2020, 23, 273–274. [Google Scholar] [CrossRef]
  23. Sozzi, M.; Marinello, F.; Pezzuolo, A.; Sartori, L. Benchmark of satellites image services for precision agricultural use. In Proceedings of the AgEng Conference, Wageningen, The Netherlands, 8–12 July 2018; pp. 8–11. [Google Scholar]
  24. Li, J.; Knapp, D.E.; Schill, S.R.; Roelfsema, C.; Phinn, S.; Silman, M.; Mascaro, J.; Asner, G.P. Adaptive bathymetry estimation for shallow coastal waters using Planet Dove satellites. Remote Sens. Environ. 2019, 232, 111302. [Google Scholar] [CrossRef]
  25. Pascual, A.; Tupinambá-Simões, F.; Guerra-Hernández, J.; Bravo, F. High-resolution planet satellite imagery and multi-temporal surveys to predict risk of tree mortality in tropical eucalypt forestry. J. Environ. Manag. 2022, 310, 114804. [Google Scholar] [CrossRef] [PubMed]
  26. Roy, D.P.; Huang, H.; Houborg, R.; Martins, V.S. A global analysis of the temporal availability of PlanetScope high spatial resolution multi-spectral imagery. Remote Sens. Environ. 2021, 264, 112586. [Google Scholar] [CrossRef]
  27. Battles, A.C.; Kolbe, J.J. Miami heat: Urban heat islands influence the thermal suitability of habitats for ectotherms. Glob. Change Biol. 2019, 25, 562–576. [Google Scholar] [CrossRef] [PubMed]
  28. Sun, Y.; Augenbroe, G. Urban heat island effect on energy application studies of office buildings. Energy Build. 2014, 77, 171–179. [Google Scholar] [CrossRef]
  29. Ju, J.; Roy, D.P. The availability of cloud-free Landsat ETM+ data over the conterminous United States and globally. Remote Sens. Environ. 2008, 112, 1196–1211. [Google Scholar] [CrossRef]
  30. Xiao, C.; Li, P.; Feng, Z.; Wu, X. Spatio-temporal differences in cloud cover of Landsat-8 OLI observations across China during 2013–2016. J. Geogr. Sci. 2018, 28, 429–444. [Google Scholar] [CrossRef]
  31. Xu, S.; Yu, T.; Xu, J.; Pan, X.; Shao, W.; Zuo, J.; Yu, Y. Monitoring and Forecasting Green Tide in the Yellow Sea Using Satellite Imagery. Remote Sens. 2023, 15, 2196. [Google Scholar] [CrossRef]
  32. Rouse, J.W.; Hass, R.H.; Schell, J.A.; Deering, D.W.; Harlan, J.C. Monitoring the Vernal Advancement and Retrogradation (Green Wave Effect) of Natural Vegetation; Final Report, RSC 1978-4; Texas A & M University: College Station, TX, USA, 1974; pp. 1–120. [Google Scholar]
  33. Gitelson, A.A.; Merzlyak, M.N. Remote estimation of chlorophyll content in higher plant leaves. Int. J. Remote Sens. 1997, 18, 2691–2697. [Google Scholar] [CrossRef]
  34. Grover, A.; Singh, R.B. Analysis of Urban Heat Island (UHI) in Relation to Normalized Difference Vegetation Index (NDVI): A Comparative Study of Delhi and Mumbai. Environments 2015, 2, 125–138. [Google Scholar] [CrossRef]
  35. Mostofi, N.; Hasanlou, M. Feature selection of various land cover indices for monitoring surface heat island in Tehran city using Landsat 8 imagery. J. Environ. Eng. Landsc. Manag. 2017, 25, 241–250. [Google Scholar] [CrossRef]
  36. Matlan, S.J.; Abdullah, S.; Alias, R.; Mukhlisin, M. Effect of working rainfall and soil water index on slope stability in Ranau, Sabah. Int. J. Civ. Eng. Technol. 2018, 9, 1331–1341. [Google Scholar]
  37. Suribabu, C.R.; Sujatha, E.R. Evaluation of moisture level using precipitation indices as a landslide triggering factor-a study of Coonoor Hill Station. Climate 2019, 7, 111. [Google Scholar] [CrossRef]
  38. Monsieurs, E.; Dewitte, O.; Demoulin, A. A susceptibility-based rainfall threshold approach for landslide occurrence. Nat. Hazards Earth Syst. Sci. 2019, 19, 775–789. [Google Scholar] [CrossRef]
  39. Segoni, S.; Rosi, A.; Lagomarsino, D.; Fanti, R.; Casagli, N. Brief communication: Using averaged soil moisture estimates to improve the performances of a regional-scale landslide early warning system. Nat. Hazards Earth Syst. Sci. 2018, 18, 807–812. [Google Scholar] [CrossRef]
  40. Zhao, Q.; Qu, Y. The Retrieval of Ground NDVI (Normalized Difference Vegetation Index) Data Consistent with Remote-Sensing Observations. Remote Sens. 2024, 16, 1212. [Google Scholar] [CrossRef]
  41. Eastman, J.R.; Sangermano, F.; Machado, E.A.; Rogan, J.; Anyamba, A. Global Trends in Seasonality of Normalized Difference Vegetation Index (NDVI), 1982–2011. Remote Sens. 2013, 5, 4799–4818. [Google Scholar] [CrossRef]
  42. Fayech, D.; Tarhouni, J. Climate variability and its effect on normalized difference vegetation index (NDVI) using remote sensing in semi-arid area. Model. Earth Syst. Environ. 2021, 7, 1667–1682. [Google Scholar] [CrossRef]
  43. Mazzarino, M.; Finn, J.T. An NDVI analysis of vegetation trends in an Andean watershed. Wetl. Ecol. Manag. 2016, 24, 623–640. [Google Scholar] [CrossRef]
  44. Shapiro, A.D.; Liu, W. Evaluating Land Surface Temperature Trends and Explanatory Variables in the Miami Metropolitan Area from 2002–2021. Geomatics 2024, 4, 1–16. [Google Scholar] [CrossRef]
  45. Xian, G.Z. Characterizing Urban Heat Islands Across 50 Major Cities in the United States; US Geological Survey: Reston, VA, USA, 2023. [Google Scholar] [CrossRef]
  46. Shi, H.; Xian, G.; Auch, R.; Gallo, K.; Zhou, Q. Urban Heat Island and Its Regional Impacts Using Remotely Sensed Thermal Data—A Review of Recent Developments and Methodology. Land 2021, 10, 867. [Google Scholar] [CrossRef]
  47. Rendana, M.; Idris, W.M.R.; Rahim, S.A.; Abdo, H.G.; Almohamad, H.; Al Dughairi, A.A.; Al-Mutiry, M. Relationships between land use types and urban heat island intensity in Hulu Langat district, Selangor, Malaysia. Ecol. Process. 2023, 12, 33. [Google Scholar] [CrossRef]
  48. Guha, S.; Govil, H. Land surface temperature and normalized difference vegetation index relationship: A seasonal study on a tropical city. SN Appl. Sci. 2020, 2, 1661. [Google Scholar] [CrossRef]
  49. Garai, S.; Khatun, M.; Singh, R.; Sharmaet, J.; Pradhan, M.; Ranjan, A.; Rahaman, S.M.; Khan, M.L.; Tiwari, S. Assessing correlation between Rainfall, normalized difference Vegetation Index (NDVI) and land surface temperature (LST) in Eastern India. Saf. Extrem. Environ. 2022, 4, 119–127. [Google Scholar] [CrossRef]
  50. Sharma, M.; Bangotra, P.; Gautam, A.S.; Gautam, S. Sensitivity of normalized difference vegetation index (NDVI) to land surface temperature, soil moisture and precipitation over district Gautam Buddh Nagar, UP, India. Stoch. Environ. Res. Risk Assess. 2022, 36, 1779–1789. [Google Scholar] [CrossRef] [PubMed]
  51. Kikon, N.; Kumar, D.; Ahmed, S.A. Quantitative assessment of land surface temperature and vegetation indices on a kilometer grid scale. Environ. Sci. Pollut. Res. 2023, 30, 107236–107258. [Google Scholar] [CrossRef] [PubMed]
  52. Hussain, S.; Raza, A.; Abdo, H.G.; Mubeen, M.; Tariq, A.; Nasim, W.; Majeed, M.; Almohamad, H.; Dughairi, A.A.A. Relation of land surface temperature with different vegetation indices using multi-temporal remote sensing data in Sahiwal region, Pakistan. Geosci. Lett. 2023, 10, 33. [Google Scholar] [CrossRef]
  53. Sharma, K.V.; Khandelwal, S.; Kaul, N. Downscaling of Coarse Resolution Land Surface Temperature Through Vegetation Indices Based Regression Models BT–Applications of Geomatics in Civil Engineering; Ghosh, J.K., da Silva, I., Eds.; Springer: Berlin/Heidelberg, Germany, 2020; pp. 625–636. [Google Scholar]
  54. Jaber, S.M. Insights About the Spatial and Temporal Characteristics of the Relationships Between Land Surface Temperature and Vegetation Abundance and Topographic Elements in Arid to Semiarid Environments. Remote Sens. Earth Syst. Sci. 2023, 6, 254–274. [Google Scholar] [CrossRef]
  55. Cui, Y.Y.; de Foy, B. Seasonal Variations of the Urban Heat Island at the Surface and the Near-Surface and Reductions due to Urban Vegetation in Mexico City. J. Appl. Meteorol. Climatol. 2012, 51, 855–868. [Google Scholar] [CrossRef]
  56. Almeida, C.R.; Teodoro, A.C.; Gonçalves, A. Study of the Urban Heat Island (UHI) Using Remote Sensing Data/Techniques: A Systematic Review. Environments 2021, 8, 105. [Google Scholar] [CrossRef]
  57. Li, J.; Sun, R.; Liu, T.; Xie, W.; Chen, L. Prediction models of urban heat island based on landscape patterns and anthropogenic heat dynamics. Landsc. Ecol. 2021, 36, 1801–1815. [Google Scholar] [CrossRef]
  58. Zhou, B.; Rybski, D.; Kropp, J.P. On the statistics of urban heat island intensity. Geophys. Res. Lett. 2013, 40, 5486–5491. [Google Scholar] [CrossRef]
  59. Wu, X.; Zhang, L.; Zang, S. Examining seasonal effect of urban heat island in a coastal city. PLoS ONE 2019, 14, e0217850. [Google Scholar] [CrossRef] [PubMed]
  60. Kedzuf, N.J.; Zuidema, P.; Mcnoldy, B. Seasonal Variability and Trends of the Miami Urban Island. Amer. Met. Soc. 2011, 98, 1–6. Available online: https://bmcnoldy.earth.miami.edu/papers/KZM2018_98AMS_ext.pdf (accessed on 11 May 2025).
  61. Hari Kandel, A.M.; Whitman, D. An analysis on the urban heat island effect using radiosonde profiles and Landsat imagery with ground meteorological data in South Florida. Int. J. Remote Sens. 2016, 37, 2313–2337. [Google Scholar] [CrossRef]
Atmosphere 16 00880 g001
Figure 1. Maps of study area: (a) contiguous USA, (b) map of Florida, (c) map of Miami showing different town boundaries.
Figure 1. Maps of study area: (a) contiguous USA, (b) map of Florida, (c) map of Miami showing different town boundaries.
Atmosphere 16 00880 g001
Atmosphere 16 00880 g002
Figure 2. Methodology used for deriving land surface temperature (LST).
Figure 2. Methodology used for deriving land surface temperature (LST).
Atmosphere 16 00880 g002
Atmosphere 16 00880 g003
Figure 3. Workflow of this study showing data analysis steps.
Figure 3. Workflow of this study showing data analysis steps.
Atmosphere 16 00880 g003
Atmosphere 16 00880 g004
Figure 4. VARI (Visible Atmospherically Resistant Index) raster of study area for winter and summer.
Figure 4. VARI (Visible Atmospherically Resistant Index) raster of study area for winter and summer.
Atmosphere 16 00880 g004
Atmosphere 16 00880 g005
Figure 5. NDRE (Normalized Difference Red Edge Index) raster of study area for winter and summer.
Figure 5. NDRE (Normalized Difference Red Edge Index) raster of study area for winter and summer.
Atmosphere 16 00880 g005
Atmosphere 16 00880 g006
Figure 6. NDVI (Normalized Difference Vegetation Index) raster of study area for winter and summer.
Figure 6. NDVI (Normalized Difference Vegetation Index) raster of study area for winter and summer.
Atmosphere 16 00880 g006
Atmosphere 16 00880 g007
Figure 7. LST (land surface temperature) raster of study area for winter and summer.
Figure 7. LST (land surface temperature) raster of study area for winter and summer.
Atmosphere 16 00880 g007
Atmosphere 16 00880 g008
Figure 8. UHI (urban heat island) raster of study area for winter and summer.
Figure 8. UHI (urban heat island) raster of study area for winter and summer.
Atmosphere 16 00880 g008
Atmosphere 16 00880 g009
Figure 9. Land surface temperature (Gi_Bin) hotspot raster of study area for winter and summer.
Figure 9. Land surface temperature (Gi_Bin) hotspot raster of study area for winter and summer.
Atmosphere 16 00880 g009
Atmosphere 16 00880 g010
Figure 10. Land cover classification of study area.
Figure 10. Land cover classification of study area.
Atmosphere 16 00880 g010
Atmosphere 16 00880 g011
Figure 11. Correlation matrix of vegetation indices versus LST for summer.
Figure 11. Correlation matrix of vegetation indices versus LST for summer.
Atmosphere 16 00880 g011
Atmosphere 16 00880 g012
Figure 12. Correlation matrix of vegetation indices versus LST for winter.
Figure 12. Correlation matrix of vegetation indices versus LST for winter.
Atmosphere 16 00880 g012
Atmosphere 16 00880 g013
Figure 13. Scatter plot showing linear regression between LST and vegetation indices for winter.
Figure 13. Scatter plot showing linear regression between LST and vegetation indices for winter.
Atmosphere 16 00880 g013
Atmosphere 16 00880 g014
Figure 14. The scatter plot showing linear regression between LST and vegetation indices for summer.
Figure 14. The scatter plot showing linear regression between LST and vegetation indices for summer.
Atmosphere 16 00880 g014
Atmosphere 16 00880 g015
Figure 15. Suitability analysis of urban heat island intensity effect for winter and summer seasons.
Figure 15. Suitability analysis of urban heat island intensity effect for winter and summer seasons.
Atmosphere 16 00880 g015
Table 1. Data utilized in this study.
Table 1. Data utilized in this study.
DataSpatial ResolutionSource of DataTypes of DataAcquisition Months
Landsat data 202330 mUSGS/EROSRasterJuly and November
PlanetScope Scene Data 20233–7 mPlanetScopeCOG (Cloud Optimized Geo Tiff) RasterJuly and November
PlanetScope data acquired for this study
July 2023 DataNovember 2023 Data
20230723_150912_76_245820231106_154948_35_2498
20230723_150909_77_24af20231106_154946_30_2498
20230723_150907_48_24af20231106_154944_24_2498
20230723_150905_18_24af20231106_154942_19_2498
20230723_150654_80_24ca20231106_154900_24_249a
20230723_150652_51_24ca20231106_154858_19_249a
20230723_150650_21_24ca20231106_154856_15_249a
Landsat data acquired for this study
Data Set IdentifierCloud Cover %WRS PathWRS Row
LC08_L2SP_015042_20231111_20231117_02_T112.78015042
LC08_L2SP_015042_20230706_20230717_02_T116.39015042
Table 2. Vegetation indices utilized in this study with formulas.
Table 2. Vegetation indices utilized in this study with formulas.
Vegetation IndicesFormula
NDVI (Normalized Difference Vegetation Index)(NIR − R)/(NIR + R) [32]
NDRE (Normalized Difference Red Edge Index)(NIR − RE)/(NIR + RE) [33]
VARI (Visible Atmospherically Resistant Index)(G − R)/(G + R − B) [33]
Table 3. Land cover classes of study area with area and count of pixels from study.
Table 3. Land cover classes of study area with area and count of pixels from study.
OBJECTIDValueCountArea (Sq km)Class Name
10798394771.86Developed
21296365626.67Green
Total Area98.53
Table 4. Confusion matrix showing accuracy of land cover classification of study area.
Table 4. Confusion matrix showing accuracy of land cover classification of study area.
OBJECTID Class ValueC_0C_1TotalU_AccuracyKappaClass Name
1.00C_048.003.0051.000.940.00Developed
2.00C_16.0043.0049.000.880.00Green
3.00Total54.0046.00100.000.000.00
4.00P_Accuracy0.890.930.000.910.00
5.00Kappa0.000.000.000.000.82
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

K C, S.; Chiluwal, A.; Magar, L.P.; Paudel, K. Investigating Urban Heat Islands in Miami, Florida, Utilizing Planet and Landsat Satellite Data. Atmosphere 2025, 16, 880. https://doi.org/10.3390/atmos16070880

AMA Style

K C S, Chiluwal A, Magar LP, Paudel K. Investigating Urban Heat Islands in Miami, Florida, Utilizing Planet and Landsat Satellite Data. Atmosphere. 2025; 16(7):880. https://doi.org/10.3390/atmos16070880

Chicago/Turabian Style

K C, Suraj, Anuj Chiluwal, Lalit Pun Magar, and Kabita Paudel. 2025. "Investigating Urban Heat Islands in Miami, Florida, Utilizing Planet and Landsat Satellite Data" Atmosphere 16, no. 7: 880. https://doi.org/10.3390/atmos16070880

APA Style

K C, S., Chiluwal, A., Magar, L. P., & Paudel, K. (2025). Investigating Urban Heat Islands in Miami, Florida, Utilizing Planet and Landsat Satellite Data. Atmosphere, 16(7), 880. https://doi.org/10.3390/atmos16070880

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.
Back to TopTop