Thermal Remote Sensing: A tool to Determine Temporal Land Surface Temperature in Hawassa City, Ethiopia


 This investigation was conducted for the estimation of the temporal land surface temperature value using thermal remote sensing of Landsat-8 (OLI) Data in Hawassa City Administration, Ethiopia. Satellite datasets of Landsat-7 (ETM+) for 22nd March 2002 and Landsat-8 (OLI) of 22nd March 2019 were taken for this study. Different algorisms were used to estimate the Normalized Difference Vegetation Index threshold from the Red and Near-Infrared band and the ground earth's surface emissivity esteem is legitimately recovered from the thermal infrared by coordinating with the outcome got from MODIS information. The land use land cover map of the city was prepared with better accuracy using the on-screen classification technique. The spatial distribution of surface temperature of the city range from 6.62°C to 22.54°C with a mean of 14.58°C and a standard deviation of 11.25 in the year of march 22nd 2002. The LST result derived from Landsat 8 for March 22nd, 2019, ranges from 11.97°C to 35.5°C with a mean of 23.735 °C and a standard deviation of 16.64. In both years the higher LST values correspond to built-up/settlement and bare/open lands of the city; whereas, lower LST values were observed in vegetation (trees/woodlot, shrubs, and grass forested) area. Urban expansion (built-up area roads, and another impervious surface), decline in vegetation levels due to deforestation and increasing population density. Increasing an evergreen tree and green space coverage, design and develop city parks and rehabilitate the existing degraded natural environments are among the recommended strategy to reduce the rate of LST.


INTRODUCTION
Land surface temperature is the surface temperature of the earth that is directly in contact with the measuring instrument (measured in Kelvin). As defined by Anandababu et al. (2008) land surface temperature is the surface temperature of the earth's crust where the heat and radiation from the sun are absorbed, reflected and refracted. Land surface temperature varies with a transition in environment and other human behaviors that forecasting precisely is difficult. Earth's surface energy balance, thermal properties of surface and atmospheric conditions affect the land surface temperature dramatically (Guo et al. 2012;Orhan et al. 2014). Urbanization is one of the most important factors for socio-economic development of the city (Dociu and Dunarintu, 2012;. Currently, urbanization has dramatically increased in greenhouse gasses and reshaped the environment, and has major climatic effects at all scales owing to the gradual transition of natural land use into an urban region (Yan et al. 2002;Yıldırım et al . 2011;Orhan et al. 2014). Unplanned urbanization is the product of rapid growth in size and speed, resulting in decentralized and scattered construction (urban expansion) (Bharath et al . 2014) and various environmental effects (Mohan et al . 2011). Land surface temperature play an important role in research on effects of urban heat island, and environmental monitoring (Lu and Weng 2006;Muhammad et al. 2018). Land surface temperature is the main factor in the measurement of a given location's maximum and lowest temperature.
Urban Heat Island ( UHI) definition and classification is usually based on spatially varying land surface temperature , due to non-homogeneity of the surface cover and other atmospheric factors (Jeevalakshmi et al . 2017). Remote sensing is useful for understanding spatio-temporal land cover changes in terms of surface radiance and emissivity data in relation to the basic physical properties (Tran et al. 2006;Orhan and Yaka 2016). Land surface temperature, calculated from remote sensing data, is used in many areas of science; such as; hydrology, agriculture, climate change, urban planning, etc. To evaluate the environmental problem it is important to obtain surface temperatures and use them in various experiments (Orhan et al. 2014).
Many researchers showed that Landsat having medium-resolution satellite is the only source of earth surface temperature in worldwide since 1972 (Amiri 2009;Guo et al. 2012;Orhan and Yaka 2016;. For Land Surface Temperature study Landsat 8 is the most important source of data (Muhammad et al., 2018). It is was launched on the 11 th of February 2013 into space with two instruments on board, the Operational Land Imager (OLI) and the Thermal Infrared Sensor (TIRS) (Anandabablunu et al. 2008;Muhammad et al. 2018).
According to Jeevalakshmi et al . ( 2017), OLI gathers data at a spatial range of 30 m, with eight bands distributed in the electromagnetic spectrum's visible and near-infrared and short-wave infrared regions, and additional panchromatic band of 15m spatial resolution. TIRS senses the TIR radiance at a spatial resolution of 100m using two bands located in the atmospheric window between 10 and 12 μm (Muhammad et al. 2018;Anandababu et al. 2008).
This study was conducted in Hawassa City Administration, Ethiopia, which is the central city of Southern region. Multitemporal thermal image series were acquired by Landsat -5 TM/8-OLI were used. The main objectives this study was estimating the temporal land surface temperature using thermal remote sensing.

Description of Study Area
This study area was carried out in Hawassa town and surrounding area. It is located in the southern Nations, Nationalities and peoples (SNNP) regional state. It is a city in Ethiopia, on the shores of Lake Hawassa in Great Rift Valley. It is found at the distance of 273 km south of Addis Ababa, the city is the most rapidly expand city in the country. Geographically, it is located on latitude and longitude of 6 0 55'N 38 0 25'E and 7 0 5'N 38 0 33'E in geographical coordinate Systems respectively. The city is one the industrial city of the country and has become the most beautiful city, tourist attractive city in the country. The city administration has an area of 157.2 Square km, but the overall study area has an area of 560 square Km. The mean annual rainfall was estimated to be 973mm. The mean annual temperature area was estimated to be 27.45 0 C in maximum temperature, 13.04 0 C in minimum temperature and 20.25 0 C the mean temperature. The monthly average maximum temperature of the area varies from 24.57 0 C to 29.99 0 C, the monthly average minimum temperature of the area varies from 10.36 0 C to 14.58 0 C and the monthly average mean temperature of the area varies from 19.32 0 C to 21.57 0 C.

Data Used
A Landsat 8 OLI image is one of the Landsat series of NASA. For this study January, February and March, 2019 (Path/Row-168/55) pertaining to the study area were used to estimate the temporal land surface temperatures. In the present study, bands 10 were used to estimate brightness temperature and bands 4 and 5 were also used to generate normalize difference vegetation index of the study area. Landsat 8 provides metadata of the bands such as thermal constant, rescaling factor value etc., which can also help to calculate land surface temperature. Landsat 8 Bands, Wavelength and Resolution are as shown in Table-1.

Methodology
Techniques of Thermal Infrared (TIR) remote sensing has been used to study urban climate and environment and especially for assessing and estimating land surface temperature patterns and their relationship with surface characteristics (Weng et al. 2004;Jeevalakshmi et al. 2017). The study utilized potentials of Semi-Automatic Classification plugin integrated with Open Source GIS package Quantum GIS for image acquisition, pre-processing, image classification and derivation of land surface temperature from land surface emissivity values. Detailed description of the methodology is outlined here (Figure 1). Image acquisition and Pre-processing cloud free Landsat a scene (Path/Row-168/55) was downloaded. The whole approaches used for the proposed work is shown in the Figure 2. This technique can only be used to process LANDSAT 8 data. In this research, Band 10 is used in this study to calculate surface temperature, and it is calculated using bands 4 and 5 to normalize the discrepancy vegetation index.

Method of Data analysis
There are several steps or analytical procedures to be followed in order to conduct LST estimation and analysis.
Conversion of DN values into Top of Atmosphere (TOA) Radiance is the first step in land surface temperature estimation.in this algorism all, the satellite data were geometrically corrected and band ten was used as an input data. In this activity radiance, rescaling factor is very important to retrieve the top of atmospheric (TOA) spectral radiance from Thermal Infra-Red Digital Numbers. In order to estimate the land surface temperature from the Landsat-8 thermal infrared band data, DN of sensors are converted to spectral radiance using the following equation USGS 2013;Jeevalakshmi et al. 2016;Ren et al. 2015;Julia 2014;Wang et al. 2015). Where, • ML represents the band specific multiplicative recalling factor • Qcal is the Band 10 image • AL is the band specific additive rescaling factor • Oi is the correction value for band 10 (  After converting DN values to at-sensor spectral radiance, the TIRS band data should be converted to Top of atmosphere brightness temperature (BT) using the thermal constants given in metadata file. The following equation is used as a tool in this algorism to convert reflectance to brightness temperature (USGS 2013). Calculating Land Surface Temperature is the final steps in this algorism. The surface temperature value of the study area can be calculated as follow equation (7 For ease of comprehension, the above derived land surface temperature s' unit was converted to degree Celsius using the relation of 0 _C equals 273.15 K.

Validation of retrieved Land Surface Temperature
Validation is required for independent assessment of accuracy and uncertainty of the derived output (Nikam et al. 2016).
As the Earth's surface temperature is determined from thermal remote sensing datasets utilizing dynamic methods and underlying hypotheses regarding atmospheric parameters, it is also important to determine its precision, which is beneficial for both consumers and engineers. (Li et coll., 2013). Normal land surface temperature products of the same region and same time span can be used to confirm the obtained surface temperature values. Yet due to the complex existence of land surface temperature, both spatially and temporarily, ground-based values cannot be obtained to crossvalidate the temperature of the collected surface. Furthermore, owing to the non-availability of observed surface temperature details, the cross-validation technique was used to confirm the recovered surface temperature of the earth; thus we used regular data sets of Landsat 8-OLI earth surface temperature products with precision within 1oK range (Weng et al . 2004) to confirm the earth surface temperature of Landsat 8 satellite image.

Results and Discussion
The present study has been conducted to analyze the effectiveness of multispectral satellite data to retrieve land surface  Table 2 and same has been shown in Figure 4.

Land Surface Temperature Analysis
In this study, Landsat 7 (ETM+) and Landsat 8 TIRS data has been utilized for estimating land surface temperature of study area. A Single channel (SC) equation-based method was used to estimate land surface temperature . Land surface temperature estimation is shown in Figure 5 respectively. The land use/land spread map (Figure 3)  construction is taking place. Whereas, the lower land surface temperature value was observed about 11.97°C in the forested area of the study area, while the built-up areas were increased to 35.5°C (Figure 3 and 5  April the temperature has shown higher value than the rest months. On the other hand, from May to September the maximum temperature becomes lower than other, particularly, in June, July August the maximum temperature has lower value, because these season manly know as a rainy season in the country as well as the study area. However, in most case of the selected years ( Figure 6)

Conclusion
This study investigated the spatial variability of land surface temperature over Hawassa city Administration located on the southern part of Ethiopia. This study examine the change in land surface temperature over different land use/land cover of the area over the period of time (17 years are the main causes for the increment of land surface temperature in the city. The outcome this research is demonstrate the role of geospatial technologies in land surface temperature estimation and this would help to assist studies on urban environment and effective urban land use planning.

Competing interests
The author declares that they have no competing interests.

Author contributions
Mikias Biazen Molla developed the overall methodology section, main analysis framework, collecting primary and secondary data and carried out the multi-criteria analysis. Finally the author read and approved the final manuscript.

Ethical approval
This article does not contain any studies with human participants or animals performed by any of the authors.

Informed consent
None