Analysis of Land Surface Temperature Distribution in Response to Land Use Land Cover Change in Agroforestry Dominated Area, Gedeo Zone, Southern Ethiopia

: This study examined LST distribution in Ethiopia's agroforestry-dominated Gedeo Zone due to Land Use Land Cover change. For 2005, 2011, 2017, and 2022, 10 m Sentinel 2A and 30 m Landsat images were used to extract and map LST and LULC distribution. The DOS1 method corrected atmospheric errors in all satellite images. LULC change was detected using SVM image classification. The study result revealed that the Agroforestry and Built-up coverage has increased by 1520 sq. km and 2600 sq. km, respectively, from 2005 to 2022. The Bare Land and Farm Land coverage decreased by 1554 sq. km and 2565 sq. km, respectively, in the same period. The LST result has shown that there has been a remarkable variation in the spatial pattern of the LST between 2005 and 2022. The average LST in Agroforestry, Bare Land, Farm Land, and Built-up area has progressively increased over the years, from 19.6°C, 26.0°C, 20.2°C, and 25.58°C in 2005 to 25°C, 32.16°C, 28.23°C, and 30.62 °C in 2022, respectively. While in 2005, the maximum recorded LST did not exceed 37.3°C, by 2022, it had increased by close to 3°C, reaching 40.6°C. The overall result revealed that the average LST in °C has increased from 2005 to 2022. From the result, it was concluded that agroforestry had contributed a lot to LST distribution. LST may not depend on the local LULC change only; other factors like urbanization and global warming could play a significant role in changing LST locally and globally.


Introduction
Land surface temperature (LST) measures the earth's surface temperature and is an essential variable in climate modeling and environmental monitoring [1], [2]. It is affected by various factors, including solar radiation, air temperature, humidity, wind speed, and surface properties such as albedo and emissivity [3]. Land Use Land Cover (LULC) changes can significantly impact LST by altering the surface properties and energy balance [4], [5]. Studies showed that converting forested areas to agricultural land results in decreased evapotranspiration and increased surface temperature due to reduced vegetation cover and increased soil exposure [6]- [8].
Agroforestry systems are becoming increasingly popular in many world regions as a sustainable approach to land use management [9]. Introducing agroforestry practices can increase vegetation cover and soil moisture retention and reduce surface temperature [8], [10]. Agroforestry systems can potentially mitigate the adverse effects of LULC changes and provide multiple benefits, including climate adaptation and mitigation, food security, and biodiversity conservation [11]. Agroforestry practice in the Gedeo zone of southern Ethiopia is known as old-aged and indigenous [10]. Gedeo agroforestry is a well-known land-use system, and it is believed to have self-sustaining and self-regulating attributes compared to other land-use systems in the area [11].
Agroforestry is a climate-smart farming system that is stronger in mitigating climate change than seasonal cropping [12] and gets people nearby to safe working areas for food security from the climate change perspective [13]. LST measures the earth's surface temperature and is an essential variable in climate modeling and environmental monitoring [14]. It is affected by various factors, including solar radiation, air temperature, humidity, wind speed, and surface properties such as albedo and emissivity [15]. Most artificial activities are the leading cause of the constantly declining vegetation cover of the earth's surface [16] and contribute to the rise of LST.
Even though several studies were conducted to observe the impact of LULC change on LST, the extent and direction of these changes remain poorly understood, and there is a need to investigate the impact of LULC on LST in the agroforestry-dominated Gedeo Zone. Understanding the impact of LULC on LST can help identify areas where agricultural productivity may be at risk and guide the development of appropriate adaptation strategies.
In comparing agroforestry systems and subsistence farming, agroforestry systems outperform subsistence farming [12]. The sale of numerous products acquired through the system could result in financial gain [13]. Fruits, nuts, timber, medicinal plants, animal feed, green manure, resins, gum, spices, and other supplementary or diversified items can only be harvested from an agroforestry system, which is especially useful for smallholder farmers [14].
It is an ancient and indigenous agroforestry practice in the Gedeo zone of southern Ethiopia [15]. A well-known land-use system is the Gedeo agroforests, which are thought to be more self-sustaining and self-regulating than other regional land-use patterns [16]. The factors that affect the diversity and composition of agroforestry in the Gedeo zone and elsewhere in Ethiopia have been studied, as have the management of indigenous agroforestry methods, the interaction of agroforestry system components, and other related issues [17], [18].
The extent and direction of these changes remain poorly understood, and there is a need to investigate the impact of LULC on LST in the agroforestry-dominated Gedeo Zone. Understanding the impact of LULC on LST can help identify areas where agricultural productivity may be at risk and guide the development of appropriate adaptation strategies.

An Explanation of the Research Site
All the research was done in Ethiopia's Gedeo zone, part of the Southern Nations Nationalities and Peoples (SNNP) region. The geographical range covered by the investigation extends from 5°53' N to 6°27' N and from 38° 8'E to 38°30'E in latitude and longitude, respectively ( Figure 1). This area is between 1,500 and 3,000 meters above sea level. Annual precipitation averages between 800 and 1800 mm, and average annual temperatures range from 12.5 to 25 degrees Celsius in this region [19].

Data Sources
Two years of Cloud-free Sentinel-2A MSI images were downloaded from European Space Agency (ESA) website for land use land cover classification in 2017 and 2022. Sentinel image has a 10 m spatial resolution relatively higher resolution than Landsat images. Three scenes of sentinel images were mosaicked and subset to the study area to address the extent of the study area. Additionally, four years of Landsat imageries with a path/row 168/56 were downloaded from the USGS-EROS website (https://eros.usgs.gov/) for land cover classification and LST extraction. A 1.50 m and 5.00 m spatial resolution Spot images and Google Earth historical images were utilized for assessing the accuracy of classification for 2005, 2011, and 2017.  [20]. Accordingly, four classification classes (Agroforestry, Bare land, Built-up, and Farmland) were identified in the study area. The SVM was preferred due to its higher performance than other classifiers [21]. One hundred representative GPS points were collected from each LULC class to assess the classification accuracy in January 2022.

Land Surface Temperature Retrieval
LST extraction from Landsat images involves converting the digital numbers (DN) values in the thermal band of Landsat imagery into LST values [22]. Landsat sensors' thermal band measures the radiation emitted from the earth's surface in the thermal infrared wavelength range [23], [24].

Conversion of DN to Radiance
The raw DN values in the thermal band of Landsat images are not directly proportional to the emitted radiance from the earth's surface [25]. The DN values were first converted to radiance using radiometric calibration coefficients provided by the USGS [23].
Utilizing the rescaling factors stored in the metadata file, the following equation was used to convert DN values to Radiance for OLI-TIRS.
where: Lλ = TOA planetary Spectral Reflectance; AL = Reflectance additive scaling factor for the band; QCAL = L1 pixel value in DN; MP = Reflectance multiplicative scaling factor for the band.

Conversion of Radiance to Sensor TB
TB was calculated by equation 1 from the spectral radiance of thermal bands [24], [26].

LULC Changes (2005 -2022)
From (Figure 3), we can observe that the area of Agroforestry increased steadily from 104,071 sq. km in 2005 to 105,591 sq. km in 2022. This indicates that agroforestry practices have successfully maintained and even expanded the forest cover while providing for agricultural needs. This suggests that the region actively engages in agroforestry practices to manage the land and maximize productivity sustainably.
On the other hand, the Bareland area decreased from 6725.7 sq. km in 2005 to 4163.13 sq. km in 2011, but it slightly increased again to 5169.78 sq. km in 2022. This suggests that efforts to reforest and restore degraded lands have been partially successful, but more work needs to be done to prevent further deforestation and land degradation. The built-up area has consistently increased, from 1092.6 sq. km in 2005 to 3693.87 sq. km in 2022. The area of Farmland has been decreasing over the years, from 36,515.7 sq. km in 2005 to 33,950.3 sq. km in 2022. This could be due to urbanization, changing agricultural practices, and land degradation.   This study's analysis of the LULC dynamics in the Gedeo Zone, dominated by agroforestry, showed a growing demand for land for urban growth. As a result, from 1092.6 sq km in 2005 to 3693.87 sq km in 2022, a larger area was covered by built-up areas. Another Ethiopian study supported this, concentrating on the need for land in Ethiopian cities and towns for urban growth [31].
The coverage of Agroforestry increased steadily from 104,071 sq. km in 2005 to 105,591 sq. km in 2022. This indicates that agroforestry practices have successfully maintained and even expanded the forest cover while providing for agricultural needs [29]. This suggests that the region actively engages in agroforestry practices to manage the land and sustainably maximize productivity [32].
The overall analysis of the LST distribution result revealed that the mean LST in °C has increased from 2005 to 2022. This could be due to urbanization, changing agricultural practices, and land degradation [33], [34]. Even though the area coverage of Agroforestry increased due to the dynamics of other LULC, other factors like global warming could play a vital role in the rise of average LST. Thus, the rise of LST may not depend only on the local LULC change [35], [36]. Other factors like urbanization and global warming could significantly change LST locally and globally [37].
Understanding the impact of LULC on LST is vital for identifying areas where agricultural productivity may be at risk. The Northern areas, including Dilla Town and its surrounding, have experienced significant urbanization, with a maximum recorded mean LST between 2002 and 2022. Understanding the impact of LULC on LST is vital for identifying areas where agricultural productivity may be at risk [38]. The Northwest and Southern parts of the study area, bounded by the West Guji zone, has dominated by Bare Land and negatively influenced agricultural productivity [39]. These areas should be targeted for interventions such as irrigation and shade management. These interventions can help to mitigate the adverse effects of high temperatures on agricultural productivity and ensure sustainable agricultural production in the region.

Conclusion
The overall dynamics of LULC in the study area indicated a gradual shift towards sustainable practices such as agroforestry, urbanization, and changes in agricultural practices. Policymakers must monitor and manage these changes to ensure sustainable land use and minimize negative environmental and livelihood impacts. From the result of the study, the following conclusions were attained: The area coverage of the built-up area and bare land has increased rapidly while the coverage of agroforestry increased slightly from 2002 to 2022, leading to the rise of LST. The increased area coverage of agroforestry may not guarantee the cooling down of LST; other factors like urbanization and global warming could play a significant role in changing LST locally and globally.