LULC Dynamics and the effects of Urban Green Spaces in cooling and mitigating micro-climate change and Urban Heat Island Effects: A case study in Addis Ababa city, Ethiopia

Land Surface Temperature (LST) increment and Urban Heat Islands (UHI) variability are the major urban climatology problems arising in the urban development. Greening urban environment play vital role to combat the effects of micro-climate change. This study attempts to assess the effects of Urban Green Spaces in cooling and mitigating micro-climate change in Addis Ababa City. Three different dates of remotely sensed data from Landsat5 TM (1990) Landsat7 ETM+ (2005) and Landsat8 OLI/TIRS (2021) were used for the study. LST were retrieved from Landsat5 TM and Landsat7 ETM+ using mono window and Landsat8 OLI/TIRS were used split window algorithms. Regression and correlation analysis of LST, Normalized Vegetation Index (NDVI) were performed in SPSS V23. The Results from this study have shown that the proportion of Urban Green Space (UGS) to other LULC were reduced from 120.4 Km 2 in 1990 to 76.26 Km 2 in 2021. The Result of multiple linear regression analysis clearly indicates that built up and green vegetation contributed 92.2% of the LST variations in Addis Ababa City. The Cooling E�ciency (CE) and Threshold value (TVoE) of green space in Addis Ababa City were calculated as 4.5 ± 0.5 ha. This �nding indicated that the city municipality implements urban planning, allocating a green space area of 4.5 ± 0.5 ha is the most e�cient to reduce heat effects of the study area. The result of the study indicates that, strengthening of public participation in urban greening as an important strategy to mitigate the effects of micro-climate change and it is important to sustain urban development as well as to provide better quality of life on the urban population.


INTRODUCTION
Microclimate is the most important element in modern global climate study as far as the world-wide climate change is a cumulative result of local microclimate's impact (Hessen, 2018).Climate change has contributed to a signi cant increase in the global mean temperature (IPCC, 2014).Earth's surface temperature is a result of the balance between incoming solar energy and outgoing radiation energy (Roza et al., 2017).Urban heat islands (UHI) is an event resulting from rapid urbanization which is described as urban areas with signi cantly warmer temperatures than its nearby rural areas (Kong, 2016).There are a number of contributing factors, which play signi cant role in creation of UHI such as low albedo materials, wind blocking, air pollutants, human gathering, and increased use of air conditioner.It is a major sample of microclimate at work (Nuruzzaman, 2015)."Rapid urbanization driven by population growth and economic development has led to drastic and widespread changes to the Earth's surface, causing the replacement of natural surfaces such as vegetation with impervious surface materials such as concrete, asphalt and buildings" (Angel et al., 2019).Increased replacements of the natural greenery area to urbanized areas have led to signi cant changes in the local climate conditions (ktri et al., 2017).
A more practical method of mitigating the UHI is strategic planting of vegetation in urban areas and designing green technology approach (Ari n, & Amin, 2013).Urban greenery also acts as a natural agent against air pollution in the urban environment (Buyadi et al., 2015).The urban planning, preparation, and implementation strategy allocated 30%, 30%, and 40% for roads any and infrastructure, green areas and shared public use, and building construction in their urban land management plan respectively (MoUDH, 2015).In Ethiopia, urbanization causes extreme changes in vegetation cover, hydrological administrations, and nearby scale climates.The most obvious climatic impact of urbanization is an increase in LST in urban areas relative to the surrounding rural areas (Feyisa et al., 2014).Most studies neglect the cooling effect of all urban green space patch for mitigating microclimate.Besides other urban parameters, built-up and green space play a signi cant role in LST dynamics (Berhanu Keno, et al.,2021).By applying geospatial technique of land use Transfer Matrix Method (LUTM), the cooling impact of green space patches on UHI mitigating is properly analyzing (Shaker et al., 2019).
Addis Ababa has for long been the capital city and a hive of economic activities in the country.The city is in a construction boom (Antos et al., 2016) and observing the scale of development and construction activities occurring on Addis Ababa, Young (2014) described the entire city as it seems one construction site.Especially in recent years, the city has witnessed a number of changes with new development and redevelopment programs-the construction of mega structures such as malls and business centers, international hotels, high ways and creation of new settlement areas.According to RUAF Foundation (2020), Addis Ababa is the thirty rst fastest growing City in the world with an annual growth rate of 3.4 from 2006-2020.
Urban LST is highly dependent on the physical or thermal properties of an object, it is vital to see the size, thickness, and dissemination of green spaces in urban regions to decide their effect on UHI and distinguish the connection with LST.Even though different researchers try to asses on green space as an approach of mitigating UHI in some cities of the country even in Addis Ababa, the method applied for these studies were not adequate to clearly determine the contribution of each green space patches in mitigating UHI.For instance; (warkaye, 2016, Feyisa, 2014, Ermiyas,2017, and Samson et al, 2018) used buffer zone setting to explore the cooling e ciency of some selected parks without considering other green spaces like greenery alongside road and road divide, road squares, cemetery, trees in churches and nonreligious institutions; and emphasis hasn't given on each green space patches., and the emphasis hasn't given on each green space patches.The quantitative role of green space is still uncertain, which limits the ability to make speci c recommendations for urban land use optimization, as well as the ability of landscape planning and design to mitigate the heating effects within a city.For example, the exact extent and intensity to which a speci c green space design would impact the wider urban thermal environment are not well studied.The Minimum -Maximum threshold of the size of green space required to provide cooling and an optimal size with the highest cooling e ciency was not studied in the study area.Moreover, there is no su cient study regarding urban morphology type and its thermal effect in Addis Ababa and the study tried to examine the thermal effect of different urban morphology types Therefore, this study aims to analyze the implication of UGS in cooling e ciency and micro-climate moderation by taking the parameters of LULC, LST, UHI, NDVI and NDBI of the study area and expected to provide an understanding of the Land Surface Temperature (LST), Urban Heat Island (UHI) and LULC status of the area as an input for planning and decision-making.

Description of the study area
Addis Ababa is the capital and the largest city that lies in the central highland of the Ethiopian federal government and on the western edge of the Rift valley in the Eastern Africa region (Dissanayake, et al., 2016).Geographically Addis Ababa is located 38°40′00″-38°52′30″E and 8°52′30″-9°50′00″N with a topographic variation range from 1,700 m to 2,000 m above sea level (Fig. 1).Based on the national metrological agency the mean annual rainfall at Kotebe, Bole, and Akaki stations are 960mm, 998mm and 955.mm respectively (Wubneh, M., 2019).
The minimum and maximum mean annual temperature in Addis Ababa about 120C and 240C, respectively (Abo-El-Wafa, H.et al., 2018).The night and day time temperature of the city is 10-15 ºC and 20-24 ºC respectively during the dry period (Ki e 2017).The Ethiopia's urban population will anticipated in triple to 42.3 million by 2037 (World Bank, 2015).Addis Ababa hosts an estimated 3.2 million people, which is a 17% share of Ethiopia's total urban population (Kidanewald, 2018).

Data sources and description
Landsat images TM of 1990, Landsat ETM + of 2005, and Landsat OLI/TIRS of 2021 were used for this study (Table 1; Fig. 2).The Landsat images were downloaded from the USGS Earth Explorer website (https://earthexplorer.usgs.gov/)which is freely available to users.All Landsat images were downloaded with little cloud cover (< 10%) and during the dry season (January, February).The Landsat images were used for LULC change analysis and calculation of LST, NDVI, and UHI.

Accuracy assessment
Accuracy aim of accuracy is to perform and evaluate the quality of classi cation output Error matrix based on assessment of the overall accuracy; producer's accuracy, user's accuracy, and kappa coe cient (Eq. 1) were utilized to evaluate the pixel-based classi cation output for LULC classi cation (Bhatta, 2008).284 Sample data from the eld using GPS for the recent period of 2021 and Digitized ROI Polygon and Google earth visual interpretation of the selected LULC type were prepared for the period 1990 and 2005.The result of an accuracy assessment provides an overall accuracy of the map based on an average of the accuracies for each class in the map (Habtamu, 2011).
Where: OAC is over all accuracy, UAC is users' accuracy, PAC is producer accuracy, Khat is Kappa statistics is total number of samples, Xij is the diagonal values, xi + is the column total, and X + i is row total, r is the number of categories, Obs = it represents accuracy reported in error matrix (overall accuracy) and Exp = it represents correct classi cation.Trends of LULC (ha) = X2 -X1 …………………Eq (2) Where: X1 is initial year, X2 is the nal year.Where: Y1 and Y2: area coverage of LULC at initial year (Y1) and nal year (Y2), Z is time interval between two years.
2.3.5 Land use and land cover change matrix.
In the present study, the LULC change detection for Addis Ababa city was applied by using the land use transfer   2004), the emissivity is calculated using (Eq.7).ε = 0.004 *PV + 0.986 …...…......Eq.( 7) Where; PV is Vegetation Proportion,  = [ − −] 2 NDVI is Normalized Difference Vegetation Index; whereas, NDVI max is Maximum Normalized Difference Vegetation Index and NDVI min is minimum Normalized Difference Vegetation Index.The calculated radiant surface temperature is corrected for emissivity using the equation (Eq.8).

LST Validation (Accuracy Assessment)
The LST estimated from the Landsat imageries were cross validated against the measured data which was received from the Ethiopian National Meteorological principal station in Addis Ababa city administration.Correlation analysis is used to cross validate the estimated and measured temperatures within the study area.Correlation function is employed to evaluate the relationship between the estimated average temperature and measured average temperature from measured weather station (Eq.10).
CC = corr (T1, T0) …………………………Eq (10) Where: CC = correlation coe cient, T1 = average temperature from meteorology station, and T0 = average temperature estimated from satellite image 2.7 Transformation of land surface temperature into Urban Heat Islands (UHI) Seasonal variation and different atmospheric conditions within the same period among years is not appropriate to compare multiple data images from different years.Therefore, to compare seasonal variation of Urban Heat Islands (UHI) from different dates, lad surface temperature normalization methods were performed using the following equation (Abutaleb et al., 2015) (Eq.11).UHI = T−T/SD…………………………..Eq.( 11) Where: Ts = land surface temperature, Tm = mean of the land surface temperature of the study area, and SD = Standard deviation.LST > µ + 0.5 * is considered an UHI area; Eq. ( 14), LST ≤ µ + 0.5 * is characterized as a non-UHI area; Where; = standard deviation of LST, and µ = mean of LST.

Driving the Green Spaces' Cooling Effect
To examine the contribution of green spaces in relation to other LULC class in mitigating UHI were performed using the following equation developed by (Choi et al., 2010).With the mean temperature within the study area as a baseline, the thermal in uence of each of the functional zones to the entire urban landscape was computed by the Contribution Index (CI) (Eq.12).

CI = D*St……………………Eq (12)
Where: CI Contribution Index -the in uence of the zone to the entire landscape, D: is the difference in mean temperatures between the zone and the entire urban landscape, and St: is the proportion of the area to the entire landscape.The green space cooling effect is determined as the difference in temperature between inside the green space and the average land surface temperature of the entire landscape (Choi et al., 2010).To determine the cooling effects of green spaces, the surface temperature of green spaces and the entire landscape should be computed (Eq.13).

Calculation of the CE
Many studies revealed that the cooling effect of urban green space patches decreases with distance from the boundary of the patch and disappears at a particular distance (Bowler et al., 2010).
The CE is de ned as a curve between the realm (Green patch) and therefore the maximum cooling intensity of every green space.If the scale of green space increases from P1 t P2, the change of ΔLST (µ2-µ1) is kind of signi cant.
However, if the scale of the green space increases from p3 to p4, there's only a really small increase in ΔLST (µ4 -µ3) while the change of area is that the same.
Accordingly, propose a TVoE during this study.As an example, the ΔLST increases with the dimensions of green space, which suggests that the CE is productive.At a speci c point, the CE will become stable within a speci c range.Therefore, now is that the TVoE.

Correlation and Regression analysis
Correlation and regression analyses between LST and NDVI associated with urban LULC type were computed using different techniques such as linear regression method by integrating with zonal statistics data to identify the correlation.The impact of the green spaces and built-up areas on LST is quantitatively described (SPSS V23, and RStudio).(Fig. 3).The LUTM method was employed to drive the quantitative description of state transition system analysis (Fig. 4).LULC matrix was produced by overlaying two LULC maps of the same area to show the probability that one particular LULC category changed into another category.In this study, from initial to nal year transitional land cover matrices were produced for every three periods of the study.The result shows each class changes from the rst column stands for the initial state of LULC categories to the second stand for the nal state of LULC categories (Table 4).Built-up was one of the lands uses classes which had shown high increment throughout the study period  The Minimum UHI variation in the last three decades were observed in the Northern part of the study area with natural and plantation forest, in the parks and around the river banks, and where riverine vegetation existed.This nding is in line with a number of studies conducted in Ethiopia (Samson, 2018; Teferi and Abraha, 2017; and Kemal, 2019).The pattern of UHI only extracted from northern to the central part of the city.The result indicates, UHI increased with high frequency from the northern part towards the center of the city.Particularly, the spatiotemporal variation of UHI in the south, southwest/east part of the city was very high due to the expansion of industries, infrastructure, commercial centers, built-up.The results also indicated that the maximum UHI increased spatially by the rate of 231.6 Km 2 from the year 1990 to 2021(Table 6).The correlation between NDVI and LST was found to be negative at all speci ed years, as shown in Fig. 7. i.e., R 2 = 0.917(91%), 0.920(92%), and 0.9291 (92%), It can be concluded that, if the area is densely vegetated, the LST is found to be lower.This correlation is said to be indirect.When the correlation is said to be negative, it refers not to the weakness of the correlation, rather it describes the decrease in the number of NDVI values, it the increase in the LST value.Trees and other plants help cool the environment, making green space a simple and effective way to mitigate urban heat island effects.Therefore, the effects of the increase in patch density on LST can be explained by a decrease in mean patch size of green spaces.Generally, negative correlation was found between NDVI values with LST.This nding is in agreement with a number of studies.For instance, (Zhou et al., 2011; Feyisa et al., 2014) were show that green spaces can lower surface and air temperatures by providing shade that prevents land surfaces from direct heating from sunlight.
The result shows that the mean NDVI value was decreased with increment of LST of the study area.As a result, the LST presented as increment order from < 18ºC to > 30ºC shows the decrement of the mean NDVI Value With 0.37 to 0.07, 0.3 to -0.02, and 0.29 to 0.13 in the year 1990, 2005, and 2021 respectively (Table 11).This study identi ed that areas covered with matured trees and other green vegetation tend to have lower temperature.Studies conducted by Asgarian et al., (2015) and Sun et al., (2013) reveal that the spatial pattern of green spaces contributed more to the cooling effect variation than the total area of green space did, and the complexity and connectivity of green spaces can enforce the stability of land surface temperature.
The decrease in mean patch size may increase LST because a larger continuous green space produces stronger cool island effects than that of several small pieces of green spaces.The results of this study showed that NDVI was correlated with LST with statistical signi cance.This nding is consistent with a number of previous studies, which demonstrated negative correlations between LST and the abundance of green space measured by the NDVI (Buyand et al., 2015; Kemal,2018).
The scatter plot computed in SPSS V23 showed a negative relationship between LST with NDVI and a positive correlation with NDBI in each of the Three years.In this research multiple linear regression analysis was applied.The model produced through this analysis shows that these two urban parameters (NDBI & NDVI) contributed to 92.1% of the LST variations with the R² = 0.92 of the study area.Thus, the value of LST in Addis Ababa City can be estimated using these two urban parameters with reasonable accuracy.This means that areas with lower vegetation cover are experiencing higher land surface temperature and vice versa.Other studies have also shown similar  Based on the model produced, for areas with the same NDBI measure, an increase in NDVI of 5% implies an expected decrement of LST by 1.4ºC.Whereas, in areas with the same NDVI, an increase of 5% in NDBI value implies an estimated increase by 1.6 ºC LST.An increase in LST can affect the human thermal comfort of the study area.
While a Mature green cover needs to be planted and maintained within an area to balance the negative effect of the built-up area.This result is in line with (Isa et al., 2017) point out that the two urban parameters NDVI and NDBI have signi cant effects on the LST; and con rmed that built-up areas have a greater in uence on the LST as compared to the green areas.
The model produced in this study (LST = 31.22X1-27.402X2+ 48.17 Mainly, the Sub city inhibited with industrial and factories zones (Akaki, Lafto and Bole), Similarly the central part of Addis Ababa (Arada, ldeta, chirkos and Addis Sun city) where observed the highest LST due to the area dominated with business center, extremely dence built up and vehicles along the road side and throughout the entire part.This study has similar result (Kemal, 2018, and Husen, 2019) that found, increment of built-up area will result in severe effect of urban heating whereas the increment of green areas seen to be the most suitable measures to reduce the LST.

The cooling e ciency of urban green space
Many studies revealed that the cooling e ciency of urban green space patches decreases with distance from the boundary of the patch and disappears at a particular distance (Bowler et al., 2010).Some researchers de ned the urban cooling effect because of the phenomenon of lower temperature within a green space patch compared with surrounding built-up areas (Lin et al., 2015).Similarly, during this research, the result shows the cooling effect of green space patches because of the LST difference between the patch and its surrounding geographic region.
Accordingly, de ne the most cooling extent because of the distance between the sting of the green space and also the rst turning point of a temperature drop compared with the green space's temperature.
In the last three decades, green infrastructure has gained popularity as an effort to increase green areas in the cities.
It is notable that the area has the highest correlation with the maximum temperature difference.In this manner, expanding the region of the green space seems successfully improve the cooling impact.Numerous studies (Feyisa et al, 2014, and Yan et al., 2018) reveal that urban parks in uence by the cooling of the surrounding areas and mitigating the urban heat.The CE is expressed as a logarithmic curve between the area of each UGS and its maximum ΔLST.Variations in LST and CE of green spaces in Addis Ababa City have evaluated.Areas of green spaces patches are in a logarithmic relationship with the maximum temperature difference (ΔLST) with the coe cient correlation of (R² =0.684).This indicates that the areas of green spaces are highly related with the cooling range and maximum surface temperature difference.Based on the statistics of the cooling range appropriate to the areas of the green spaces, a regression analysis of the area with the cooling extent and maximum surface temperature difference was made, respectively (Fig. 8).The nding from this study is also in agreement with a number of studies such as Kong et al., (2016), and Lai et al., (2019) that green spaces have a potential in uence on the thermal environment.
The turning point of the curve is the maximum Land Surface Temperature change (ΔLST) and de ned as the cooling intensity of urban green spaces.The ndings of this study to calculate the cooling extent and intensity of each green space patch.The range of 4.52 ± 0.5ha can bring a maximum change in surface temperature (ΔLST).This indicates that the bigger the size of green space, the higher the CE.The gradient of the model curve is very steep on the lefthand side which means that the CE of green space increase with the size of green spaces sharply.On the right-hand side of Fig. 20, the gradient is low and the CE has become reduced and stable.A threshold value of the CE of green space is then calculated as (TVoE = 4.52 ha).The formula of the CE curve of green space is: y = 0.6009ln(x) + 2.0002 (R² = 0.684).Furthermore, to further validate the results, the size of green space patches was divided into ve segments.In between the patch, which has very small in size and closest to each was added for driving accurate result and further validate the above discussion (< 1 ha, 1-2 ha, 2-4.5 ha, 4.5-15 ha and greater than 15 ha (Fig. 9).
The result indicates that in the 0-4.5ha segments, correlations between ΔLST and the size of green space show a positive relationship, which changes to a negative relationship in the range > ) that shows urban space were more prone to UHII due to the high share of built-up area and industrial area whereas, UHII was comparatively low due to the presence of greenery.Not only built-up but also fallow land and barren land contributed signi cantly as heat source (Table 10).illustrates that vegetation has a great impact on reducing LST.The strong negative correlation between NDVI and LST denotes that the higher the vegetation a land covers has, the lower the LST.Because of this relationship between NDVI and LST changes, Land Use Land Covers have an indirect impact on surface temperatures through NDVI.Thus, it is very crucial to promote and support urban greening, such as new planting in the public areas and green infrastructure.Systematically, the scatter of green spaces should be uniform with a total area of 4.5 ± 0.5ha to obtain the optimum cooling effect in Addis Ababa City.Increasing the green area within 4.5 ± 0.5ha can improve its cooling effect with minimum economic cost.Since this study has veri ed the value of remotely sensed data in understanding the in uence of urban LULC types particularly green spaces on LST; the quantitative evidence from remotely sensed data can be valuable in designing current and future sustainable urban socio-economic and environmental plans.To address the urban green cooling effect in the face of decreasing urban vegetation, key strategies include increasing vegetation through tree planting and green spaces, preserving existing vegetation, implementing green infrastructure like green roofs and walls, promoting urban agriculture, integrating vegetation into land use planning, providing incentives, raising awareness, and fostering collaboration among stakeholders.
These measures contribute to mitigating the urban heat island effect and creating cooler, more sustainable cities.

CONCLUSIONS
The advancement of Remote Sensing and Geographic Information system become convincing towards analyzing the environment we are living today.In this study, Landsat data for the year 1990(TM), for 2005 Landsat ETM + and for 2021 Landsat 8(OLI were used to generate important information on LULC; and to interpret the relationship of LST with NDVI and NDBI for the determination of the cooling e ciency of green spaces in mitigating urban microclimate change in Addis Ababa City, Ethiopia. The study showed that, there is the rapid expansion of the built-up areas and deterioration of vegetation cover and agricultural land was the most important deviations which could be a major possible cause for the UHI effect in Addis Ababa City.NDVI of the year 1990 having the value > 0.3 was reduced from 65.06 km 2 (12.35%) to 29.The study also revealed that there is a signi cant land surface temperature difference among the urban morphology types.Based on the eld observation and percent land cover analysis, villa has the lowest surface temperature, which can partly be explained by the existence of green structure in the area than the other UMT, while Apartment is experienced with high LST distribution than the others UMT.The role of green vegetation in urban climate was also analyzed.Differences in temperature between the areas covered with vegetation and their surroundings clearly reveal that vegetation can minimize surface temperature i.e., the increasing green space size can signi cantly mitigate UHI effects.Addis Ababa is a Metropolitan city in Ethiopia; issues of LST and UHI are very sensitive due to the expansion of the built environment, deterioration of vegetation cover, and small area coverage of the urban green spaces, which is getting worse these days.Therefore, urban planners and decision-makers were calculated as 4.52 ± 0.5ha which means that when the city municipality implements landscape planning, a green space size of 4.52 ± 0.5ha is the most e cient to cool the heat effect.

DECLARATIONS CONFLICT OF INTEREST
The writers claim there isn't any con ict.Map of LULC conversion from the speci ed study years).
Page 26/30 Results of the CE of green spaces in Addis Ababa city The graph shows the Correlations between area sizes and ΔLST Supplementary Files

2. 3 . 3
Trends and rate of LULC change Trends of LULC classes were the main indicator of how much LULC types had changed in the study area over time (Moisa et al. 2023).Trends of LULC types of the study area were calculated by using the nal and initial year of LULC classes.The greater values of nal LULC classes were related to positive trends and vice versa (Eq.2)

2. 5
Normalized Difference Vegetation Index (NDVI) estimation NDVI is used for monitoring the growth and health of vegetation as well as to detect any stress or damage.In addition to mapping and categorizing different vegetation types, NDVI values can be used to evaluate changes in vegetation cover over time (Halder et al. 2022; Zhu et al. 2022).Consequently, it has been used to calculate the abundance of vegetation cover on the Earth's surface (Alam et al. 2022).The NDVI value was calculated by using multispectral bands from Landsat images taken in 1990, 2005, and 2021.Bands 4 and 3 of Landsat 5 and 7 were utilized to measure NIR and red, respectively.Furthermore, in Landsat 8, bands 5 and 4 were employed to assess NIR and red, respectively (Eq.9).NIR = the pixel digital number of TM Band 4 and Band 5 for Landsat7 and 8 respectively, and R = DN of TM Band 3 and Band 4 for Landsat7 and 8 respectively (Gandhi, 2015).
and there was small change into others class in the study years from 1990 to 2021.In this study, the results show the two largest percentage share of land conversion from one type to another are vegetation 62.2Km2 (49.84%), 25.7Km2 (33.2%), 64.05 km2(51.31%)and farm land 52.92Km2 (23.20%), 82.91 Km2 (70.5%), 160.11Km2 (70.2%) mainly in to built-up areas and some extent into others land use class in the years from 1990 to 2005, 2005 to 2021 and 1990 to 2021 respectively (Patra et al., 2018).
) was used to estimate the value of LST of Each and selected hot spot sub city.The Expansion of built-up (NDBI) and Green Vegetation (NDVI) lie between 5%-25% leads the results of estimated LST of Addis Ababa will become extreme by increment of more than 1.6 o C − 6 o C while the green vegetation (NDVI) will be the most suitable measures to reduce the LST value more than 1.4 o C to 5.3 o C.

39km 2 ( 5 .
58%) in the year 2021.It clearly shows that expansion of built-up land has caused signi cant land cover change as well as changes in the LST.In this regard, LST distribution in Addis Ababa is very closely related to the distribution of vegetation cover (NDVI) and built-up areas (NDBI) with the value of R2 = 0.93, and R2 = 0.92 respectively.

Figures Figure 1
Figures

Figure 5 a
Figure 5

Table 2 :
Description of LULC and UMT of study area matrix from 1990 to 2021.The rate change, losses and gain of LULC within the study period were calculated by using the change matrix in the study area (Pal & Ziaul 2017; Zhang et al. 2021).
2.7.1 Variation of UHI in Response to Land Use Land Cover Change Equation developed by (Ma et al., 2010, Senanayake et al., 2013, Effat et al., 2014) to determine where urban heat is characterized critical areas based on LST distribution and available vegetation.

Table 3
Area of actual LULC and changes in the year from 1990 to 2021.Bare Land (BL), LU; Land use Type, Built-Up (BU), Farm Land (FL), Vegetation (VC), * Negative, positive sign (-, +) shows the declining and increment trend of the class respectively.Among the four major LULC classes, built-up area rank 1st in terms of the size and percentage increment from 1990 to 2021.It is rapidly increasing with corresponds to the context of rapid urbanization in Addis Ababa City.
(39.03%).which is the most signi cant change in land cover type proportion.Dagnachew, (2018) also pointed out that built-up area consumes a considerable amount of land from vegetation cover and farm land during urban development.The overall classi cation exactness of the LULC precision appraisals for 1990, 2005, and 2021 was 93.2%, and 94.6%, 96.6%respectively.As a result, for the think about periods 1990, 2005, and 2021, the kappa coe cients were 0.89, 0.92, and 0.97.The pattern of land use land cover in the study area showed a signi cant change during the investigation period (1990 to 2021).

Table 4 LULCC
Matrix from the year 1990 to 2005, 2005 to 2021 and 1990 to 2021 Result of Land Surface Temperature, NDVI and NDBI The minimum temperature of the study years 1990, 2005 and 2021 were accounts 12.42 ºC, 14.96 ºC, and 16.15 ºC respectively.This temperature mainly observed in the Northern part of the study area where the vegetation cover is high, the mountain with natural forest (Entoto, Yerer, and others), in the parks and around the river banks, where riverine vegetation existed Fig. 9.The result shows the maximum temperature of 32.58 ºC, 38.56 ºC, 39.49 ºC were recorded in the years 1990, 2005, and 2021 respectively.The highest rate of urban expansion was observed south and west direction.Factors accelerating the increase in this direction were suitability of land, accessibility of transport and others demographic factors.These areas are mostly dominated by both government and private led housing program such as condominium, real-estate and others constriction program like road, industries, and factories zone leads high LST increment throughout the study period.The spatiotemporal trend of Land Surface Temperature variation was clearly identi ed from the map in Fig.5.In 1990 the large extent of Addis Ababa was dominated by temperature between 23 _26ºC, in 2005.Several of researches conducted on analyzing UHI, depend on generating LST, which shows the relative warmth of cities According to Teferi and Abraha, (2017) in Addis Ababa the LST average value of Built-up was increased from the period of 1986 to 2011 than other type of land use land cover signatures.The average surface temperature of Addis Ababa City has been increased from 24.68 ºC in 1990 to 28.28ºC in 2005 and 30.25ºC in 2021 at the rate of 1.8°C per decade.The study clearly reveals that an increment of surface temperature is mainly because of the decrease of green spaces that replaced by impervious surfaces.It was found that 213.5 km 2 (38.49%) of Addis Ababa City was covered by built-up areas in the year 2005, and 333.74 km 2 (60.15%) in the year 2021(Table5).More than half percent of the city was packed with man-made features which made the temperature to be higher than the surrounding regions.This result is in agreement with Varshney, (2013) that the distinctive land surface temperature patterns are associated with the thermal characteristics of land cover classes.Nobre et al., 2016, and Samson et al., 2018 also showed the evidence for climate change including the occurrence of drought, rising temperature, ood, reduced annual rainfall, and rising sea levels that have resulted from LULC dynamics.
Dagnachew, (2018) also pointed out that built-up area consumes a considerable amount of land from vegetation cover during urban development.A study conducted by Mulugeta et al., (2017) revealed that there is a high rate of urban expansion in peri-urban area of Addis Ababa and this urban expansion is the major factor for land use land cover change of the area and the case for declining of farmland and vegetation.Leulsegged et al., (2011) also noticed that Addis Ababa urban expansion was due to high demand of house in the city and massive housing program construction activities are listed as the cause for the decline for agricultural land in the city.3.3 through measuring the air temperature, using land-based observation stations (Ashraf, 2015).Weng et al. (2004), executed research in Indianapolis shows that the trend of hot spot areas has directly related with continual urban expansion.Likewise, the result of this research shows, Maximum LST expansion were observed in most parts of Addis Ababa due to rapid urban expansion in all directions except northern escarpment bounded by mountain.

Table 6
Gandhi, 2015,Gandhi, 2015.The values of NDVI can further be categorized as non-vegetated for the value < 0, unhealthy vegetation for the value close 0.03 and moderate values represent bush and grasses were ranging to 0.3, whereas, high NDVI value correspond to dense vegetation close and greater than 0.6.

Table 7
Area of actual NDVI Distribution and changes in the year from 1990 to 2021 As indicated in Fig. 6 andTable 7 vegetation cover has decreased and the non-vegetated area has been increasing gradually over the study period.Mainly by comparing MIN & MAX NDVI Value of the three different periods (1990, 2005, and 2021), it is observed that the minimum NDVI Value (non-vegetated & Unhealthy) increase from 165.08 km 2 (31.3%), 252.6 km 2 (47.94%), and 326.11 km 2 (61.89%) in the year 1990, 2005 and 2021 respectively.While the maximum NDVI values were decreased from 65.06 km 2 (12.35%), 35.946 km 2 (6.82%), and 29.39 km 2 (5.58%) of the study area was covered by the same NDVI class in 1990, 2005, and 2021 respectively.This result has been con rmed with a number of studies that shown NDVI values of dense vegetation, river banks and around water bodies experience higher NDVI values than other classes, owing to the presence of agricultural land (Feyisa et al., 2014; Samson et al., 2018).

Table 9
Show Correlations coe cients of LST, NDVI and NDBI 4.5ha segment.This result demonstrates that the calculated TVoE in this study is reliable.The TVoE of the study clearly shows that smaller green spaces have a positive relation with ΔLST, but green space with an area greater than 4.52 ± 0.5ha has a negative relation with LST change.Some studies also investigated that the cooling effect of green spaces had a size-related threshold value.For instance, Yu et al., (2020) were found out that 4.95ha of green space area is a consistence threshold value for cooling.Yu et al., (2018); Zhaowu, (2018) also indicated that 4.55ha of green space area is signi cantly important threshold value for cooling and agreed that the maximum cooling extent of UGS is expressed as the distance between the edge of the vegetation cover and the rst turning point of temperature drop compared with the UGS temperature.Heat contribution is evaluated by computing the average temperature values for the City's Green space in relation to other land use/cover types.In the four LULC types, only vegetation cover had negative heat contribution within the study area (-2.62ºC) in the year 2021.While, built-up areas and bare land (vacant spaces) had a positive contribution index.Consequently, built-up areas can be considered as the main heat source (4.28°C) followed by bare land (vacant spaces) which accounts for (+ 1.7ºC).Marginally, dry farm land had also a positive heat contribution (0.23ºC) in the year 2021 of the study area.This result is in line with different studies were conducted plenty of researchers(Odindi andMutanga, 2015, Li et al., 2017, Mukherjee et al., 2017, Rasul et al., 2018

Table 10
Based on this result, it can be assumed that the heat source/sink role in a built-up/vegetation cover are the vital role for UHI variation.Vegetation cover can therefore be regarded as the most valuable heat sink whereas built-up can be considered as the main heat source within the study area.This result agrees with numerous studies that have been conducted on the relationship between UHI and urban LULC change (Mahmood et al., 2010, Lu et al., 2015, Firozjaei et al., 2018, Pramanik et al., 2019).Find LULC with urban micro-climate patterns is crucial to understand and mitigate the impacts of urban LULCC on local micro-climate change.Urban greenspace planning plays a crucial role in improving the quality of human settlements and the living standard of citizens.Urban public greenspace (UPGS) is an important part of urban greenspaces (Min et al., 2021).The relationship between LST and NDVI clearly