Urbanization is the intensification of inner-city inhabitants and urban land cover expansion due to the relocation of people into urban regions and the transformation of rural spaces into urban areas. By 2050, it is projected that 68% of the world’s population will reside in urban areas, with global urban land area expected to continue expanding rapidly ((Nations, 2018); (Chen et al., 2020)). The rise in land surface temperature is primarily driven by unplanned rapid urban development. Urban expansion and built-up areas contribute to an increase in land surface temperature (Rahaman et al., 2023). Therefore, the urban population will be exposed to climate change risks from the rise in urban land surface temperature (Vohra et al., 2024).

In today’s world, climate change is one of the most pressing challenges we face. Land surface temperature (LST) is a crucial factor in understanding climate change and its effects (Li et al., 2022). It is also a significant parameter for a variety of applications, such as agriculture, hydrology, and urban planning (Ghasempour et al., 2023). In the past, LST was estimated using interpolation after the calculation of particular sample points. However, with the advent of earth observation satellites, LST is now routinely estimated using remote sensing data, which offers global spatial coverage, long-term records, and high spatiotemporal resolution (Zhang et al., 2021). Thus, LST can be obtained and used easily for environmental monitoring and management. Landsat 9 was successfully launched in September 2021. Like Landsat 8, it supersedes previous Landsat generations in both radiometric and geometric measurements, in addition to its high imaging capacity (N.A.S.A., 2021). Moreover, Landsat 9 confirmed greater accuracy in land surface temperature analysis in addition to being equipped with advanced OLI and TIRS sensors compared to Landsat 8 and other previous Landsat satellites (Ghasempour et al., 2023).

When compared to other satellite systems with thermal sensors, such as MODIS (Moderate Resolution Imaging Spectroradiometer) and Sentinel-3’s SLSTR (Sea and Land Surface Temperature Radiometer), Landsat 9 stands out for its higher spatial resolution. While MODIS offers daily global coverage with a spatial resolution of 1 km for thermal bands ((Zhang et al., 2021); (N.A.S.A., 2022)), Landsat 9 provides finer thermal imagery at 100 m resolution, making it more suitable for localized studies. On the other hand, Sentinel-3, with a resolution of 1km in thermal bands (E.S.A., 2013), surpasses Landsat 9 in terms of frequent revisits but lacks the spatial detail provided by Landsat 9. This balance of high spatial resolution and improved radiometric accuracy positions Landsat 9 as a crucial tool for detailed surface temperature studies, complementing broader datasets from other satellites.

Various studies have explored LST from diverse perspectives globally and to some extent in Ethiopia. For instance, (Fikriya et al., 2022) investigated the impact of built-up areas on urban LST using Landsat data in Surakarta, Indonesia, revealing the significant influence of urban structures on temperature variations. Similarly, (Li et al., 2022) examined the relationship between urban expansion and the urban heat island effect in Surakarta, highlighting a positive correlation between urbanization and LST. In the context of Africa, (Li et al., 2022) explored urbanization and climate change impacts on thermal urban environments, emphasizing the challenges posed by urban heat islands in rapidly urbanizing areas.

In Ethiopia, (Yeneneh et al., 2023) studied the relationship between land use/land cover (LULC) changes and LST in the northwestern highlands, demonstrating how changes in vegetation and urban development influence thermal patterns. Similarly, (Worku et al., 2021) analyzed the effects of vegetation cover change on LST in Addis Ababa, underscoring the critical role of green spaces in regulating urban temperatures. Earlier research by (Damtie, 2017) focused on land surface change estimation in Bahir Dar city, providing valuable insights into urban expansion and its thermal implications.

This study distinguishes itself from previous research by its method, objectives, and the agroecological zone of the study area, which focuses on estimating LST in and around urban towns, incorporating both urban and surrounding rural environments. This approach allows for a comprehensive understanding of the interactions between urbanization and LST beyond urban boundaries. For instance, the research methodology combines the emissivity equation model (Rahaman et al., 2023) with NDVI-based emissivity correction, addressing limitations of prior studies in Ethiopia that used simpler split-window algorithms (Damtie, 2017).

Furthermore, the study area encompasses two distinct climatic regions of Ethiopia, the highlands (Dega) and dry humid regions, enabling an analysis of the relationship between NDVI (Normalized Difference Vegetation Index) and LST across varying climatic conditions. By addressing these unique aspects, this research contributes to the broader understanding of LST dynamics in diverse agro-ecological zones.

This research aims to estimate the LST of the study area and examine the relationship between NDVI and LST using Landsat 9 satellite images for the year 2023, focusing on Injibara town and the surrounding Banja district, Ethiopia. This region was chosen due to its sensitivity to climate variability, being a predominantly agricultural area where temperature changes directly impact crop productivity and local livelihoods. Moreover, rapid urbanization in Injibara town has raised concerns about urban heat island effects and their implications for regional climate and human comfort. By enhancing the understanding of LST in and around the research area, this study aims to support climate change studies and provide a basis for developing sustainable land management and urban planning strategies. Additionally, this research seeks to address the knowledge gap on current LST patterns in the region. It also aims to understand the relationship between LST and NDVI in different Ethiopian climate regions, serve as a reference for similar studies, and inspire further research on climate change impacts in Ethiopia. This paper is structured as follows: Section two reviews LST retrieval techniques and regional studies; Section three details the study area and methodology; Section four presents NDVI and LST results; Section five discusses spatial patterns and their implications; and Section six concludes with policy recommendations and future research directions.

This paper is structured as follows: Section two reviews LST retrieval techniques and regional studies; Section three details the study area and methodology; Section four presents NDVI and LST results; Section five discusses spatial patterns and their implications; and Section six concludes with policy recommendations and future research directions.

This section presents the main results of the estimated Land Surface Temperature and Normalized Difference Vegetation Index, along with their correlation, in Injibara town and the surrounding Banja district. Additionally, these results will discuss the values of NDVI and LST compared to other literature.

The retrieval of LST from Landsat 9 thermal bands reveals a wide temperature range from 14°C to 43°C across Injibara town and the surrounding Banja District. This variation reflects the diverse land cover, elevation, and climatic conditions in the study area. Spatial analysis shows a clear correlation between LST and vegetation cover, as indicated by the NDVI patterns. Low vegetation areas, particularly in the southern and southeastern regions (e.g., latitude 11° 3' 49", longitude 36° 38' 00'), exhibit the highest surface temperatures (38-43°C), dominated by barren land. This is consistent with recent findings, which highlight that barren and sparsely vegetated areas typically experience the highest temperatures due to reduced evapotranspiration and higher surface albedo. For example, a study in Rajkot, India, found that regions with barren land consistently exhibited higher LST, while areas with increased vegetation cover had lower surface temperatures. Similarly, recent analyses in Ernakulum District and other regions confirm a strong negative correlation between NDVI and LST, with vegetation cover acting as a key moderator of surface temperature extremes (Chen et al., 2022).

Conversely, the cooler temperatures (14–22°C) observed in highland areas with denser vegetation (e.g., latitude 10° 57' 09", longitude 36° 53' 09" in the northwest) underscore the influence of elevation and vegetation in mitigating surface heat. However, notable variations within the same NDVI range suggest that topography plays a critical role in LST distribution. For instance, at latitude 10° 57' 34" and longitude 36° 56' 05" (center of the study area) and latitude 10° 46' 46" and longitude 37° 3' 10" (southeast), areas with similar NDVI values exhibit contrasting LST ranges due to their differences in elevation. This highlights the complex interplay between vegetation, elevation, and land cover in shaping LST dynamics in Injibara and Banja District.

Urban areas in Injibara town reveal unique thermal characteristics compared to rural surroundings. Despite low vegetation cover, built-up areas in highland regions exhibited lower surface temperatures compared to lowland regions with barren or sparsely vegetated land. This underscores the role of elevation in moderating urban heat island effects, even in semi-arid settings. However, the moderately high temperatures (29-32°C) prevalent across approximately 40% of the study area, particularly around latitude 10° 54' 21" and longitude 36° 54' 21", reflect the pervasive influence of semi-arid conditions and human activities, such as urban expansion and land degradation.

The regression analysis (Y = -27.34x + 36.42, R² = 0.2506) confirms a statistically significant inverse relationship between NDVI and LST. The R² value indicates that approximately 25% of the variation in LST can be explained by the variation in NDVI. Higher NDVI values (0.4-0.5) correspond to lower temperatures (20-24°C), whereas areas with NDVI values below 0.2 experience higher temperatures (25-41°C). This suggests that while vegetation cover plays a role in influencing surface temperatures, other factors, such as elevation, soil moisture, and anthropogenic heat, also contribute to LST patterns in the study area. This finding aligns with previous studies that have reported a negative correlation between NDVI and LST (Maharani et al., 2021)(Rahaman et al., 2023), further validating the role of vegetation in moderating surface temperatures. However, this study also reveals the importance of considering topographic variations, which can lead to deviations from this general trend.

Interestingly, the study also identifies instances where the inverse relationship between NDVI and LST is not strictly observed. Specifically, we found that some highland areas with low vegetation cover exhibit lower LST compared to lowland areas with relatively higher vegetation cover. This apparent anomaly underscores the significant influence of topography on thermal dynamics in the region. Higher elevations generally experience lower temperatures due to decreased air pressure and increased adiabatic cooling (Ma et al., 2022). Recent studies confirm that elevation can be a dominant factor in controlling LST, sometimes overriding the effects of vegetation cover. For example, (Rahaman et al., 2023) and (Ma et al., 2022) both report that in mountainous or highland regions, LST is often more strongly correlated with elevation than with NDVI, due to the cooling effects associated with higher altitudes. Similarly, (Chen et al., 2023) demonstrate that topographic variation significantly modulates LST patterns in complex terrain, highlighting that elevation-induced cooling can outweigh vegetation effects in specific contexts. These findings align with our results and emphasize the need to consider topography when interpreting LST-NDVI relationships in heterogeneous landscapes.

Injibara and Banja District face significant challenges due to extreme temperatures in lowland areas, raising concerns about the vulnerability of these zones to land degradation and climate change. These findings have critical implications for urban and rural planning. For Injibara, urban planners should prioritize integrating green infrastructure, such as urban forests and vegetative buffers, to mitigate urban heat island effects and enhance thermal comfort for residents. In the surrounding Banja District, afforestation and community-led forest conservation efforts are vital to reduce surface temperatures and combat land degradation.

The limitations of this research, which focuses on LST estimation from Landsat imagery for a single year, underscore the need for further studies. Temporal analysis over multiple years would provide a broader understanding of climatic and anthropogenic influences on LST and NDVI dynamics. Integrating socioeconomic data, such as population growth, land use changes, and economic activities, could offer valuable insights into the human dimensions of land degradation and temperature variation in the region. Moreover, refining the LST retrieval methodology to incorporate field-based validation would enhance the accuracy and applicability of remote sensing data for local decision-making.

This study demonstrates the potential of Landsat 9 data to effectively analyze LST and its drivers in diverse landscapes, such as Injibara town and Banja District. The findings highlight the critical role of vegetation, elevation, and urban planning in regulating surface temperatures and emphasize the need for sustainable land management strategies to address the challenges posed by rising temperatures and climate change.

This study successfully retrieved LST from Landsat 9 data using emissivity equations, calculated NDVI, and analyzed their correlation. The findings revealed that high vegetation areas, with NDVI values ranging from 0.4 to 0.5, exhibited low surface temperatures between 20°C and 24°C, while low vegetation areas, with NDVI values below 0.2, experienced higher surface temperatures ranging from 25°C to 41°C. The regression analysis (Y=-27.34x+36.42, R 2 =0.2506) confirmed an inverse correlation between NDVI and LST in the same topography, where areas with higher NDVI consistently corresponded to lower LST values. To support regional climate resilience, we recommend integrating green infrastructure in urban planning and prioritizing afforestation in Banja's lowlands. Future research should focus on expanding the analysis to other Ethiopian agro-ecological zones, incorporating higher spatial resolution thermal data sources where available, to validate and refine these findings across diverse environmental conditions.

Acknowledgements

The authors would like to express their sincere gratitude to all parties who have contributed to the research and the completion of this manuscript. Special thanks are extended to mentors, funding institutions, and data providers for their invaluable support and assistance throughout the study.

Author Contributions

Conceptualization: Kassahun, M.T., & Workineh, N; methodology: Kassahun, M.T., & Workineh, N; investigation: Kassahun, M.T., & Workineh, N; writing—original draft preparation: Kassahun, M.T., & Workineh, N; writing—review and editing: Kassahun, M.T., & Workineh, N; visualization: Kassahun, M.T., & Workineh, N. All authors have read and agreed to the published version of the manuscript.

Conflict of interest

All authors declare that they have no conflicts of interest.

Data availability

Data is available upon Request.

Funding

This research received no external funding.