Analysis of Surface Temperatures Distribution over Paris - Research Paper Example

Analysis of Surface Temperatures Distribution over Paris Student Name: Contents Contents 2 Abstract 3 1.Introduction 3 1.1Background 3 1.2Objectives 4 1.3Study questions 5 2.Methodology, Results and analysis 5 2.1Introduction 5 2.2Conversion of raster bands from Digital numbers (DN) to Reflectance and AT Satellite Temperature 7 2.3Land cover classification of the study area 8 2.4Reclassification to emissivity values 9 2.5Conversion from At-Satellite Temperature to Land surface temperature 12 3.Challenges faced 14 4.Discussion 14 5.Conclusion 15 References 16 Appendix 17 Abstract This study seeks to determine the ability to use GIS softwares in satellite dasta analysis. The study uses the QGIS for the task of determining the Land surface temperature profile of Paris. The analysis of data involved 4 steps namely: the conversion of raster bands from digital Number to reflectance and At satellite, land cover classification, reclassification of the land cover classification to emissivity and then conversion from At satellite temperature to Land surface Temperature. The study shows the ability to analyse the spatial using QGIS. This emanates from the successful analysis of Paris surface temperatures. The results indicate there is differential spatial distribution of temperature in Paris. From the study, it can also be concluded that Paris experiences UHI phenomenon due to differential temperatures between adjacent locations. 1. Introduction 1.1 Background Surface temperature of any given place is a very important component to life. The temperatures determine the weather pattern of a region (Lydolph). Paris has been an area of interest for temperature studies due to observed heat waves and related temperature related effects including other weather elements such as wind, which depend on the surface heat properties (Santamouris). Determination of the general surface temperature distribution can be hard to find using the surface measurements such as spatial recording. However, satellite information can give a detailed surface temperature representation. This is achieved based on the fact that satellite mapping takes a snapshot of the spatial characteristics of a given place and stores it for later retrieval and study (Lyon and McCarthy). Bearing in mind that the general surface temperature distribution of Paris is of interest due to the observed phenomena related to thermal characteristics, the study undertook to investigate the Land surface temperature distribution over the area. This could only be achieved through analysis of the satellite information about the area. Only this source has the ability to give real surface information including, but limited to temperature. By studying the spatial surface temperature of the region, it is easy to conclude and hence find possible solution (which may be long term) to the problems associated with temperature. Usually the raw data as provided by the satellite can be very useless since nothing can be concluded from it. It must therefore be analyzed to produce maps that be interpreted on a visual basis (i.e, visually useful). One such a tool for analyzing the data is the Quantum GIS. This is a desktop geographic information system (GIS) application which enables easy data viewing, editing, and analysis into something visually meaningful (Lyon and McCarthy). The software is known to be among the best GIS softwares with which users are able t create maps with multiple layers using different map projections. Based on the ease of use and improved accuracy and reliability (Westra), the software was adopted for this study. Before describing how this software was used in analysis of the surface temperature, it is important to note the objectives of the study. 1.2 Objectives Pursuant to the study, the main objective of this study is to assess the ability to use the GIS software in analyzing data into a visually meaningful and correct display. In pursuit of the main objective and based on the specific case study, there are specific objectives that guided the study in attainment of the main study. The specific objective in this case is to determine the spatial distribution of the surface temperatures over Paris (France). Alongside this, the other objective is to determine whether Paris has the urban heat island characteristics. These objectives were guided by the study questions as stated below. 1.3 Study questions Is the Land Surface temperature of Paris uniform? Does Paris display the UHI characteristics? Answering these two questions helps fulfilling the two specific objectives. A procedurally meaningful result of the specific objectives helps in determining the ability to QGIS in data analysis. Finding answers to the questions was determined by the method followed as shown in the next section. 2. Methodology, Results and analysis 2.1 Introduction As a measure of the ability to produce accurate results within the context of GIS analysis, a comparison with analysis results from another source was considered a good step. The study selected is the one carried out by European Space Agency (2013) for comparison. The selection was based on the fact that this was a similar study carried on the same area and therefore expected that the difference between the results should be small. It was also considered that as a reputable institution, the outcome would be more accurate due to the editorial peer review factor (Weller). First of all, it is important to understand the surface characteristics if the study area. Paris is mainly covered by the urban surfaces, vegetation and agricultural fields. It is also good to note that infrared band is the most useful when it comes to the assessment of the differential spatial distribution of surface temperature. As such, it can therefore, be used in studying the urban heat island phenomenon. The study was carried out based on the same principle. This study uses the semi-automatic approach to estimate the land surface temperature by use of the thermal band and land cover classification. The study uses QGIS plugin known as the Semi-Automatic Classification. The plugin allows the automatic conversion of the Landsat thermal band to AT-satellite Brightness Temperature which is then converted to land surface temperature through. Before commencement of the procedure, it is important to understand the predetermined steps that were noted down to be followed in this section. The tasks to be accomplished in this study, as put forward by Mesev included the following in the order in which they were undertaken. 1. Conversion of raster bands from Digital Numbers to reflectance and At Satellite Temperature 2. Land cover classification of study area 3. Reclassification of the land cover classification to emissivity values 4. Conversion from At Satellite Temperature to Land Surface Temperature The tasks were carried out in the order because they have a stepwise interdependency with each other in the order in which they were executed. First, the image was downloaded. It is available from the U.SA Geological survey. Based on the ease of accessibility, this is a 2013 image. The data can be found in the appendix table 6. As such, this study would mainly apply to this year. However, the study is still valid in fulfillment of the main objective. The image, being a subset of the entire scene, included the following Landsat bands: Blue, red3-Green, Near-infrared Short wavelength, infrared 1, Short wavelength infrared 2 and Thermal infrared 1 (TIRS 1). Each of the bands is a single 16 bit raster. The stepwise conversion to Land surface temperature proceeded as shown in the sub-sections below and as outlined about the procedure above. 2.2 Conversion of raster bands from Digital numbers (DN) to Reflectance and AT Satellite Temperature The process began with the conversion of the raster bands from DN to reflectance. Then, At Satellite Temperature. The DN is converted to the physical measure of the Top Of Atmosphee reflectance (TOA) and the thermal band to At-Satellite Brightness Temperature. Simple atmospheric correction was applied to the bands from 2 to 7 (the image-based DOS1 method). In order to this, the QGIS was opened and the semi-automatic classification plugin started. Then, Landsat was selected from the Preprocessing tab. The director containing Landsat bands including the required metafile were selected from the desktop and the output directory selected where the converted bands were saved. DOS1 atmospheric correction was then applied and the whole process of conversion from Landsat to reflectance done using the perform conversion button. The process saw all the converted bands getting loaded in QGIS as shown in the picture below. 2.3 Land cover classification of the study area This process is necessary in defining the land surface emissivity. The surface emissivity is a mandatory step in converting to Land surface Temperature from AT-Satellite brightness temperature. First the land surface cover was identified into four classes which include soil, vegetation, built-up and water. The stud performed the land cover classification process by defining the classification inputs (Landsat bands from 2 to 7) and then creating ROIs. This process can be visualized in figure 5 in the appendix. The study used the spectral angle mapping algorithm to classify the study area. Below is a workflow description of the whole process up to the classification process. A detailed description of about the supervised classification process was relied on for guidance can be found at (QGIS, 2014). The classification process produced an image as shown in the image below. 2.4 Reclassification to emissivity values The stud did not carry out field survey to find out on the emissivity values, the conventional indicative emissivity (e) values were used in reclassification of the land cover classification to emissivity values. He values used are as shown below Table 1: classification values In the QGIS, the following steps were followed; the tool Reclassify grid values was selected from the processing toolbox as shown using the path, SAGA > Grid - Tools > Reclassify grid values and then in the tool window, classification.tif was selected from grid and then under method, [2] simple table was selected. The lookup table button was then clicked to open a window Fixed Table; There are 3 columns including new, minimum and maximum. The minimum and maximum define the reclassification range for new value as sown in the table below Table 2: Reclassified data The emissivity raster was then saved in the folder emissivit raster on the desktop folder parisproject. Jus for a view, the interface looked was as below. After sme seconds, the reclassified grid (emissivity) was loaded in the QGIS as shown in the image below. 2.5 Conversion from At-Satellite Temperature to Land surface temperature On the processing toolbox of the QGIS, SAGA > Grid - Calculus > Raster calculator was followed and then on Raster layers, the emissivity raster and the brightness temperatures were selected. The emissivity raster was above brightness raster. Under the fomula the equation below was written b / ( 1 + ( 10.8 * b / 14380 ) * ln(a) ) where a is the emissivity raster and b is the brightness temperature raster. This is the bass of having emissivity above brightness temperature. A few seconds later, the Land Surface Temperature in Kelvin was displayed as in the diagram below Figure 1: Land surface temperature (Kelvin). The analytical results from ESA analysis were as shown below. Figure 2: Land surface temperature for Paris 2013 (ESA 2013) 3. Challenges faced Despite the little experience gained during the classwork, the analysis of data seemed quite a task. It therefore involved lots of research both online and in books. This consumed a lot of time in the analysis. This lack of sufficient experience threatened the results of the analysis. However, this was resolved by offsetting time for quality. One biggest challenge about the project outcome is that the temperature given on the surface gives an assertive outlook of the spatial distribution. It cannot be concluded with certainty on the discrepancy size. Thus the conclusions made about the temperature difference are based only on the theoretical expression of the anticipated temperature impact. This is based on the fact that the computed temperatures are an approximation. Being a continuous variable, therefore it cannot be accurately and scientifically correct to conclude about the temperature gradient and associated impact between two points. In order to have an intrinsic understanding, the real computed, rather than color coded vales, are used for verifying the temperature difference. 4. Discussion The surface temperature map produced by the analysis of the Paris and the environs, indicates varying surface temperatures over the region of study. From the large contrasts between the adjacent areas, it can be concluded that Paris experiences internal heat island phenomenon (UHI) for some of the regions due to the identified surface temperature discrepancies between adjacent areas as can be seen in the ma displayed above. To back up this achievement, a comparison of the results between this project and the peer reviewed project show, to the highest level of similarity, that this project was successful in achieving the main objective. 5. Conclusion This study managed to determine the surface temperature distribution map of Paris without procedural challenges. From the surface temperature distribution map conclusions about the surface thermal characterization of the city can be drawn. As such, the objective to determine the surface temperature distribution was fulfilled. As the main objective of this study, it can be concluded that the study shows ability to develop and analyze maps using the GIS software (QGIS). The ability helped in finding the answer to whether there’s uniform temperature distribution over Paris. References Lydolph, Paul E. The Climate Of The Earth. Totowa, N.J.: Roman & Allanheld, 1985. Print. Lyon, J. G and Jack McCarthy. Wetland And Environmental Applications Of GIS. Boca Raton: CRC Press, 1995. Print. Mesev, Victor. Integration Of GIS And Remote Sensing. Chichester, West Sussex: Wiley, 2007. Print. QGIS, Estimation. "From GIS To Remote Sensing". Fromgistors.blogspot.com. N.p., 2014. Web. 23 Apr. 2016. Santamouris, Matheos. Cooling The Cities = Rafraîchir Les Villes. Paris: Presses de l'Ecole des mines, 2004. Print. Weller, Ann C. Editorial Peer Review. Medford, N.J.: Information Today, 2001. Print. Westra, Erik. Building Mapping Applications With QGIS. Packt Publishing, 2014. Print. Appendix Read More