SSD Analyze for The Assessment of Long Term Thermal Effect of 1 Urbanization on The Izmir City’s Local Climate Change.

7 Background: Urbanization provides several opportunities to human being to live better and comfortable life. On 8 the other hand, it also comes with some costs and side effects like worsen climate conditions. In local concept, 9 pre-climate conditions in rural area can be called as natural when they are compared against post-climate 10 conditions after urbanization expands over and swallows these natural areas. So, these natural conditions are 11 changed to worsen conditions by some civic activities in cities through urbanization. One of the urbanization 12 side effects is thermal pollution caused by specific urban activities and patterns on land surfaces in cities. Thus, 13 thermal pollution changes city’s local climate and negatively affects the city’s comfort level at least locally. 14 There are several researches focusing on that issue in cities. Each one made its contribution to the area to build 15 up a strong knowledge. One great contribution comes from the researches focusing on analyzing time serious 16 thermal data with continuous distribution over cities. Method: Here in this research is introduced and suggested a Simulated Single Data (SSD) statistical analyze 18 method for the studies based on time serious data. Therefore the method was applied to Remote Sensing (RS) 19 LANDSAT satellites’ bands especially to time series’ thermal bands of Izmir city to reveal where generally 20 Urban Hot Spots (UHS) appear and Urban Heat Islands (UHI) develop in the city w.r.t. this SSD image from 21 long period of time. Stereo representation of the study region is also used to visually examine the topographical 22 effect on UHI distribution. Conclusions: The study clearly demonstrated that industrial regions and roads with large surfaces, somehow bare lands even with spare bushes or grassy lands and more significantly the slope urban land parts within special aspects are the main contributors of UHSs and UHI developments in the city even w.r.t. long term data. Thus those contributors affect the city pre-natural climate conditions negatively and then let UHSs to appear and UHIs to develop at and around where these urban land cover structures are located or seen in the city. Those city parts are the most risky zones that city authorities take serious actions for caring their city chronical climate 29 (thermal) conditions and to focus on for returning these zones back to their pre-natural climate environmental 30 conditions. There are also some nature based solutions that are given and suggested in the conclusion section of 31 the paper for compensation of the effects caused by those contributors in the city. 32

Islands (UHI)" (Kim, 1992). According to the United State Environmental Protection Agency (US EPA), the term of "Heat Island" describes "built up areas that are hotter than nearby rural areas and the annual mean air to construct a strong, reliable and sustainable decision support system to meet their smart cities' requirements at high levels.

345
As they are mentioned previously, here can also be summarized as those; types of buildup and impervious 346 areas like buildings, concrete, asphalt structures and some specific urban activity areas (like those areas where 347 industrial activities are) in cities are the main causes of urban heat islands developments which effect the wide 348 areas in the city and then change the local climate at these city parts (Gartland, 2008). Transformation of natural 349 lands into pavements, buildings and other urban infrastructures even decreases natural cooling in cities (Tsoka et 350 al., 2020). Also, regional city building structure and pattern with multi-storey adjoining buildings and with 351 narrow streets can heat the air trapped between those buildings and also reduce airflow as mentioned above

404
Even if urban forestry is widely recognized and practiced in developed countries and also it is less known in 405 developing countries, it also offers a nature based solutions to city authorities for tussling with and mitigating 406 UHI effect in cities under high thermal pollution risk (Buyadi et al., 2013). The temperature of urban parks is 407 found to be 1-2 •C, and sometimes even 5-7 •C cooler than surrounding urban areas according to Vidrih and role to mitigate the UHI effects especially by regulating high temperature in densely urbanized areas and their

417
The current strategies to minimize the UHI effects in cities in the world are urban greening, the use of highalbedo building material, the use of suitable pavement material and proper distribution of urban buildings and

462
The aim of this study consequently became the evaluation of spatiotemporal distribution of urban heat islands 463 (UHISs) in the city of Izmir, Turkey. The study follows these steps: 1) to map emissivity for LST computation, it 464 is required to map distribution of urban green areas and the green area changes for spatiotemporal data analyses 465 by using temporarily produced NDVI images from RS MS time series' images; 2) LST spatiotemporal 466 distribution pattern are obtain across the entire city by using the RS thermal data; 3) trend images individually         10 were utilized for remote sensing and GIS data analysis and to obtain the final outputs for the entire study area.

531
Here in this study, the spatial distribution of emissivity to reveal and to correct its effect on the LST values 532 distributed all over the entire city was obtained by help of NDVI distribution for the city of Izmir having Eigen 533 climate conditions which are similar to Mediterranean humid climate conditions, since the city of Izmir is 534 located on the Eigen sea coast of the western Turkey (Fiqure 1).

535
First phase of this project was on demonstrating of LST distribution across the city of Izmir using only one 536 day data (LANDSAT multispectral image data) and representing which land use types contribute to UHI 537 development in the city. Even if one date data was used, our previous research's results emphasized that some

10/38
by data of a long period of time. Therefore, temporal remote sensing data is obtained and analyzed to reveal 542 whether the results show up with a similar output contributing our previous study outcomes thorough a certain 543 period of time. Thus we will have a robust argument to make strong and locationally precise suggestions to our 544 city authorities who will then take further actions by improving our city future plans taking into account such 545 effective and reliable arguments to be confirmed by our research outcomes to be being here in the example of 546 Izmir city. Then we would have chance to prevent our city from heat pollution and make the city much more 547 sustainable, comfortable and livable one and even to cut our cooling bills further down in the future. By the 548 decrease in energy consumption for such cooling purposes, it means low greenhouse and carbon foot print 549 effects and the saved nature or increased environmental conditions of our city by given chance to nature based 550 solutions. Thus, here in the current state of the research it is not specifically focused on only UHSs, but also 551 investigating the city regions under heavy UHI developments in a long period of time to find out which specific 552 city activities and urban land cover types cause strong and significant UHI effects. Afterward, we will have a 553 chance to take effective and correctly positioned actions to prevent our city nature in a sustainable way from    image also represents a unique UHI distribution over the city for that period of time. Thus, this project plan will 570 also help us to understand UHI distribution and its relation with the distribution of urbanized land cover types in 571 a certain long period of time. Thus, it helps us to reveal the main suspicious urban factors which strongly effect 572 the UHI developments in an urban environment even by the support of a long period of temporal data. So this 573 will encourage city authorities to take most effective precautions beforehand they happen in the future.

579
Izmir is the third of the first three metropolitan cities in Turkey with almost 4.37 million citizens living in the12 580 studied central districts which they hold the densest urban population in Izmir. The City is located on Menderes 581 and Gediz deltas and on some hilly terrain inlands almost all around the Izmir Gulf of Aegean Sea. Geographic 582 coordinates of Izmir are between 37° 49' and 39° 23' North latitudes and 26° 13' and 28° 29' East longitudes.

583
The studied city land in the province of Izmir covers almost 400 km2 urbanized area (the area covered by red 584 curved boundary in Fig. 1) around the Gulf of Izmir and the wide of the urbanized city part generally extend 3 to 585 5 km towards inlands all around the gulf. But at somewhere it reaches up to 10 -12 km, especially at the Sought 586 where the surrounding parts around the ancient city center (now called Konak) are. On the other hand, it shrinks 587 down to few kilometers at the North section of the city (where Karsiyaka district is and city newly developing 588 areas towards to the North are). This is probably because of local mountains running towards the Gulf. At those 589 parts of the city, in generally speaking, city stops extending at the forested lands where they begin to cover the 590 field and toward the tops of these mountainy areas because of the topography which rises rapidly, but not at low 591 altitude hilly lands. So, several hills next to the coast of Izmir Gulf are covered by city urban structures and 592 buildings. City also extents towards the valleys between these mountains at least at two locations more than 10 593 km inlands of these valleys, the one is seen as city sprawl at the East part and the second is similar to the first but 594 with a wider sprawl at the South. The sprawl extending towards the north is not a sprawl extending into valley, 595 this city extension only follows the Gediz delta parts just right, next to the mountain slopes running into that 596 plain delta which are not smooth hilly terrains. Another, but very narrow urban sprawl with more vegetation 597 cover can be seen at the South. This is because of rough and high mountain slopes with forest covers running 598 along very closely, right next to the gulf coast. Other very narrow sprawl is also seen in the east valley formed by high mountain slopes running towards Aegean Sea and ends up at the city center on the large plain coast formed 600 by Gediz delta. All those descriptions for the urban area of Izmir city can be followed by the stereo illustration 601 represented in Fig. 6a.      1  lt05_l1tp_181033_19850825_20171212  17  lt05_l1tp_181033_20010821_20180501  2  lt05_l1tp_181033_19860828_20170216  18  lt05_l1tp_181033_20030726_20161205  3  lt05_l1tp_181033_19870831_20170211  19  lt05_l1tp_181033_20040813_20161130  4  lt05_l1tp_181033_19880801_20171208  20  lt05_l1tp_181033_20050816_20161124  5  lt05_l1tp_181033_19890820_20170808  21  lt05_l1tp_181033_20060819_20161119  6  lt05_l1tp_181033_19900823_20171208  22  lt05_l1tp_181033_20070822_20161111  7  lt05_l1tp_181033_19910826_20171214  23  lt05_l1tp_181033_20080824_20161029  8  lt05_l1tp_181033_19920828_20180210  24  lt05_l1tp_181033_20090827_20161021  9  lt05_l1tp_181033_19930815_20180210  25  lt05_l1tp_181033_20100830_20161013  10  lt05_l1tp_181033_19940818_20180302  26  lt05_l1tp_181033_20110817_20161008  11  lt05_l1tp_181033_19950805_20180210  27  lc08_l1tp_181033_20130822_20170502  12  lt05_l1tp_181033_19960823_20180210  28  lc08_l1tp_181033_20140825_20170420  13  lt05_l1tp_181033_19970826_20180210  29  lc08_l1tp_181033_20150828_20170405  14  lt05_l1tp_181033_19980829_20170908  30  lc08_l1tp_181033_20160830_20170321  15  lt05_l1tp_181033_19990816_20180210  31  lc08_l1tp_181033_20170801_20170811  16  lt05_l1tp_181033_20000818_20161214  32  lc08_l1tp_181033_20180820_20180829 it is STAG-Sampled Tiniest Area on the Ground). Therefore, the top of atmosphere radiance does not include 682 only the radiance from an object on land surface, but also the radiance from the atmosphere along the path 683 between this instantly scanned tiniest land part which it corresponds with a certain pixel in band images in a MS 684 image scene and the sensor. Therefore, the atmospheric radiance must be removed from the total radiance 14/38

724
M : multiplicative rescaling factor for conversion from DN to reflectance (that can be obtained from the 725 metadata -REFLECTANCE_MULT_BAND and 2 x 10-5 for LANDSAT 8's reflectance bands).

726
A : additive rescaling factor for conversion from DN to reflectance (that can be obtained from the metadata -

891
After producing time series emissivity images by following the instructions given in Table 2

969
After analyses of the results as SSI output of LST distributions to reveal the hot spots and heat island 970 developments in the urbanized areas of Izmir city over the years which these urbanized regions can be followed 971 with the delineated red boundary polygon in Fig. 6a presented as a 3D illustration, it is recognized that some commercially active areas are, but none is seen at the industrial regions. So, the research pointed out that even 998 the city has 7 industrial sites ( Fig. 5 and 6a) none of them appears in these cool sites. This is a significant result 999 even if it represents only a SSI of LST distribution it is actually an output from 32 years of time series

1000
LANDSAT thermal data analyses (Fig. 6b and 6c and Fig. 7). Thus, urban parts where industrial activities are in 1001 the city almost entirely contribute to and coincide with heat island developments in the city. So, generally these 1002 sites cause hot spots to appear first and then accumulation heat problem which ends up with heat pollution as 1003 heat island and finally distribution of this heat pollution towards neighboring urban areas next to these industrial 1004 sites in the city and effect these neighboring zones in great extents (in some cases up to 5-10 km) (Fig. 6a and 6b and Fig. 7). This outcome also confirms our previous research outcome, even if it reviled the heat island

1026
Fig. 9 also represents these subsections in detail with related letter tags A to K as shown in Fig. 8. Every 1027 tagged part seen in the Fig. 9 shows details of the urban structures at that specific urbanized city parts in RGB 1028 color image form along with thematic representation of detailed SSI LST distribution related to that specific city 1029 part. Fig. 9 shows the details of every city urban subsection with corresponding tag given in Fig. 8 as mentioned   1030 above and also highlights some areas with specific urban structures in these subsections by marking them with 23/38 corresponding SSI LST distributions are also shown in the Fig. 9 next to that tagged RGB illustrations of these people who lives and works in these regions and also increases the cooling cost for bringing back the comfort 24/38

1116
Other relatively cool sites in the city are seen at residential and commercial areas delineated with large black 1117 solid line ellipse in G tagged RGB image and large black solid line circle in K tagged RGB image in the Fig. 9 1118 and they are also located on (in general) almost flat or slightly rough terrains as shown in 3D illustrations with 1119 dotted line and solid line ellipses in both stereo images of Fig. 10. It is probably because of city building 1120 structures and building materials being different than industrial sites since they are generally dwelling houses or 1121 apartments or mostly commercial buildings in city centers in these regions. So, all those work together and 1122 behave like scattering surfaces w.r.t. the sunlight coming form an angle. Therefore, these regions within described structural form do not absorb much energy but scatter it around contrary to the sites causing UHI developments. There are also some dotted line arrows which they point some sites appearing in darker blue color in both stereo illustrations of SSI LST embedded images in Fig. 10. So, these areas are the parks with mature and 1126 tall trees with large canopy surrounded by urban structures as it can be followed in the Fig. 11 as well which

26/38
represents SSI-NDVI distribution over the city of Izmir. Other cool sites marked by black arrows again in Fig.   1128 10 are slope lands facing towards North, West or North-West directions. They are the coolest areas in the region.

1154
This outcome becomes meaningful when it is compared with the similar apartment sites with tall buildings in 1155 again discrete pattern but with recreation areas between the apartment blocks decorated as car parks with asphalt 1156 or paved lands or grassy lands with some bushes and even if they are located just next to the mentioned coolest site, they appear as two of the hot spot areas in the city and they are shown with two black dotted arrows in the As it can be seen from the Fig. 13, relative temperature differences for the entire city computed appearing in dark blue color and are marked whit dotted arrows in Fig. 10) and on slopes facing towards north, 1186 west or northwest directions ( Fig. 6c and 13).  (Fig. 5, Fig. 6a and Fig. 11). High values in a NDVI image is first highly dependent on existence of vegetation

1247
There is another outcome of this study which is about building structure and building site design pattern.

1248
Depending on the SSI-LST analyses here, a dwelling site in Izmir represents quite a cool region. A certain layout 1249 pattern of this site consists of several apartment blocks in a discrete order with large common areas between the 1250 buildings. These common areas include not paved surfaces (usually which are not encountered in most cases) but 1251 green spaces with mature trees which are tall and having large canopy. Other sites with similar layout pattern

1252
(even next to this site in Izmir) but with paved surfaces between building blocks as car parking areas appear as 1253 one of the contributors of UHI development in the region (Fig. 12). In addition to that, the residential areas 1254 supported with commercial activities and buildings and also dwelling units (as apartment blocks, houses and 1255 etc.) generally show relatively cool local climates being contrary to the situations encountered in the regions 1256 under severe UHI pressures and with suspicious urban structures mentioned earlier since these cool sites include 1257 discrete or even row housing buildings and apartment blocks with low-rise storey and generally with tile roofs 1258 and most importantly they are recreationally supported by mature and large canopy trees which are closely 1259 planted around and just next to the buildings and also at the sides of the streets between these dwelling 1260 apartments and houses in the region. This mature and large canopy tree supported recreational area and housing 1261 urbanization layout style works together with the heights of low-rise and high-rise storey buildings (no matter 1262 they are for dwelling or commercial purposes) and closely planted mature trees for creating significant amount 1263 of shadowy regions in these urban sites which are generally located on almost slightly rough terrains in the city (areas with blueish colors and marked by ellipses seen in Fig. 10).

1266
Suggestions for the mitigating the UHI impact on city local climate

1268
As an essence from this study, one can come up with a conclusion such as that, if there would be no industrial

1318
As it is discussed earlier, LST shows a negative strong correlation with NDVI (Fig. 11). Moreover, most of 1319 the UHIs even from SSI analyzes are found within none vegetated areas in the Izmir city urban lands appearing 1320 as ecologically stressed zones when low vegetated areas seen as low NDVI values in Fig. 11 were compared 1321 with the areas in high LST values in Fig. 7. The natural vegetated areas such as forest and agricultural areas 1322 found at and around the boundary of urbanized lands of Izmir city as seen in the SSI NDVI image (Fig. 11) 1323 appear with low radiant temperatures in the SSI LST image (Fig. 7), so even these long period of data used in 1324 this study (compare the areas in Fig. 7 and Fig. 11). Dense vegetation can prevent lands to store high amount of