hierarchical clustering

Earth surface temperature changes can affect various environmental factors and elements, and by affecting bioclimatic comfort, energy consumption, plant and animal phenology, air pollutant concentration, soil and air moisture density, and other environmental factors can also be affected by these factors.In this research, in order to investigate the role of actual evaporation-transpiration on the temperature of the earth's surface at ten meters, the variables of carbon monoxide, water vapor density, nitrogen dioxide, sulfur dioxide, ozone, water vapor pressure and wind speed in twelve uses With the help of LCZ algorithm, determined for Tehran and Alborz provinces, have been studied and investigated.The studied area has a high potential for the occurrence of thermal islands due to special geographical conditions such as location, variety of man-made structures and special ecological conditions. In this research, according to the impact of the studied variables and the need to examine the relationships between them, hierarchical clustering in a combined manner, path analysis method and remote sensing techniques have been used. According to the obtained results, among the studied variables, actual evaporation-transpiration in none of the uses had a significant effect on increasing or decreasing the temperature of the ground surface at 10 meters. Water vapor pressure in class 14 (areas with grassy vegetation) shows the greatest effect in increasing the temperature of the earth's surface.Based on the obtained results, the type of use is highly dependent on the impact of the studied variables on the surface temperature of the earth.

IntroductionChange detection (CD) from remote sensing image considered an important topic among scientist because of its application in monitoring urban and non-urban area, environmental issue, damage assessment, etc. Presenting an efficient CD method from a high-resolution image can face a different challenge; most of the CD method from a high-resolution image requires training procedure to overcome this challenge. In this paper, an unsupervised (without needing training process) CD algorithm proposed from the high-resolution image. In this method spatial and spectral features extracted from bi-temporal images of the studied area. Difference images generated from high information content features. Then generated different images mapped into spherical space. The Primary change map created using implemented multi-thresholding method on created spherical space and the second change map created using hierarchical clustering regularized by Markov random field method. The final change map created by integrating the result of primary and secondary change maps. The final change map shows an overall accuracy of 92.56% in the studied area. Data and methodsThe data used in this paper is a subset of the main data with dimensions of 2000 * 2000 from an urban area in the city of Mashhad. These images corresponded to the two periods of 1390 and 1395 and were taken with UAV. The orthoimage is related to the first time with a spatial resolution of 6 cm and the second image is taken with a pixel size of 10 cm. In this paper, in order to detect of change of high-resolution images, first, the input images are registered in terms of spectral and spatial, and then feature images are extracted from each input image separately. In the next step, the differences images corresponding to high information content feature images are calculated. . The optimal difference images are mapped to the spherical space using selected statistical methods and in order to better analysis of the results. Otsu multi-thresholding method implemented on r component of sphere space. In the next step, the optimal difference image mapped to a spherical space is divided into non-overlapping blocks with the same dimensions; a cumulative hierarchical clustering method is applied for each block separately. In this case, the computational volume and space proposed in the hierarchical clustering method are reduced. The results of the cumulative clustering of the blocks are merged together and then the Markov random field method is used in order to regularize the results of the cluster in order to reduce noise.In final clustering, the class values below the lowest Otsu threshold are known as unchanged pixels with high reliability and the values above the maximum threshold are determined as changed pixels. The class of middle interval is unknown. For determining, the class of middle interval the corresponded output of hierarchy clustering regularized with a random Markov field is used. In the last step, a vegetation and shadow mask is used for final post-processing. Results and discussionIn order to an accurate assessment of the proposed method on the mentioned study area, a ground truth image with 11073 pixels has been used as a ground test image. The proposed method has shown an overall accuracy of 92.56 in the study area. The accuracy of detecting changed pixels shows 81.61% and the accuracy of detection unchanged pixels shows 92.77%. The false alarm percentage is 0.21 percent and the missed alarm accuracy is 0.0723 percent. For comparative evaluation, the proposed method is compared with the change vector analysis algorithm. In this section, the selected features in the feature extraction section are entered in the change analysis algorithm, and then the multi thresholding algorithm and shadow analysis used to create the final change map. This method has shown increasing the alarm in comparison with the proposed method. The accuracy of changed and un-changed pixels in the change vector analysis method is equal to 52.98 and 89.24%, respectively. Comparing these results with the results of the proposed method shows the efficiency of the proposed method. ConclusionIn this paper, the new unsupervised change detection method presented based on the combination of multi thresholding and the hierarchical clustering algorithm. Compared to supervised methods that require training data, this method does not require training data. In this method, textural and spatial-spectral features are extracted from images with high spatial resolution, which covers the discussion of the importance of neighborhoods in images with high spatial resolution. In the next step, the extracted features that have a high information content are selected, which helps to reduce the redundancy of the information. The contrast images of the features with high information content are created to differentiate the location of the changes. Spherical computing space is considered as the basic computing space. In order to create a binary change map, two analyzes have been performed on the spherical computational space. First, the Otsu multi-thresholding method has been applied. The values of the smaller and larger thresholds have definite classes. But the value of the middle interval needs to be further analyzed using the hierarchical clustering method. In this section, the middle pixel class is examined, and then a final adjustment is performed using Markov field and shadow and vegetation analysis in order to post-process and prevent false changes. In this paper, the parameters of changed accuracy – unchanged accuracy - overall accuracy - false and missed alarms have been used to evaluate the accuracy of the proposed method with a ground accuracy map. In order to make a comparative study, the proposed method is compared with the change vector analysis method of the created feature space. The results show the efficiency of the proposed method.

Precipitation is the most important hydrological variable that connects the link between atmosphere and surface processes. Three-quarters of the atmosphere heat is the result of releasing the latent heat of evaporation. Therefore, climatic zoning of autumn precipitation is necessary for achieving comprehensive development in different spatial-temporal dimensions. For this purpose, the daily data of autumn season at 18 synoptic stations of the study area have been analyzed with a joint statistical period 32 (1989 - 2018). In this study, Kolmogorov-Smirnov test was used to find out the normality of the data and Mann-Kendall test was used for the process of their changes. Then by applying factor analysis based on principal component analysis and with orthogonal rotation of Varimax it was found that there are two effective precipitation factors in the climate of the region which in total account for more than 68.16% of the variance of the region's autumn precipitation climate. By performing hierarchical cluster analysis and by integrating the matrix of factor scores, two main and sub-regions were identified. These areas are: Areas of high precipitation and moderate precipitation. It can be said that in the high precipitation areas of the territory are located on the rainy of the Zagros, Sahand and Arasbaran mountain belts, stretching as a narrow strip from the north to the south parts of the region.

According to the World Health Organization (World Health Organization, 2014), one eighth of global deaths is because of exposure to air pollution. This is more than twice the previous estimates and confirm that air pollution is currently the greatest threat to the health of the urban environment in the world.Particulate matter is one of the major air pollutants. These particles have sizes ranging from 0.000002 to 500 microns, whereas particles smaller than 2.5 microns have significant health and health effects because they are capable of accessing the respiratory system of the human body.The purpose of this paper is to use statistical methods such as cluster analysis to determine homogeneous time intervals of Particulate matter as the main pollutants of Mashhad. The results can be used by urban planners and decision makers in the discussion of air pollution management, especially planning for the control of air pollution and reducing air pollution.

Concerning the use of statistical methods in air pollution discussion Saksena Joshi and Patil (2002) examined the spatial patterns of air pollution change in New Delhi by cluster analysis method and concluded that there was no statistically significant relationship between pollutant concentration and station type. In another study, Ojeda, Cortina-Januchs, Barrón-Adame, Quintanilla-Domínguez, Hernandez, Vega-Corona, Ruelas and Andina (2010) applied PFCM to clustering algorithm to investigate air pollution in Salamanca, Mexico. Also, air pollution in Bangkok, Thailand, was analyzed using cluster analysis. Using a Geographic Information System (GIS), they mapped air pollution in the area (Prapassornpitaya, Jinsart, & Sanguansintukul 2011).To reduce costs and better management of air quality stations in the Kanto region of Japan, Iizuka, Shirato, Mizukoshi, Noguchi, Yamasaki and Yanagisawa (2014) performed cluster analysis on air pollution data and found that NOX and O3 pollutants can be classified into eight clusters. Each cluster has homogeneity in terms of spatial characteristics and temporal behavior.Ismaili (2018) investigated homogeneous areas of Mashhad air pollution in terms of particulate matter load over a two year period using GIS and cluster analysis method. Results of this study identified three separate homogenous regions for Mashhad.

In the present study, for the temporal analysis of Mashhad air, 10 air quality stations located in the city of Mashhad were used. Hourly data of suspended particles smaller than 2.5 microns (PM 2.5) were collected and arranged in matrices with temporal and spatial scales. Then a hierarchical cluster analysis was performed. Euclidean distance method and Ward method were used to calculate the degree of similarity. The resulting groups were then shown as a dendrogram.

The results of cluster analysis on PM2.5 data of Mashhad showed that three clusters or homogenous groups were recognized. The first cluster is distributed throughout the year. The second cluster with a frequency of 113 days is more mid-year. But the third cluster, which represents the highest concentration of suspended particles, often occurs in the second half of the year, with a frequency of 34 days per year.But on a monthly scale, Shahrivar, Aban and Azar are in the first cluster, and spring months and Mehr in the second cluster. Finally, Tir and Mordad and winter were clustered to the third group.Weekdays are divided into two main groups, with Friday alone in one cluster and other days of the week in the other group. At a lower level, four homogeneous clusters can be identified. Friday and Sunday are in the first and second clusters, respectively, Tuesday and Thursday in the third cluster and in the fourth cluster are Saturday, Monday, and Wednesday.

In this study, particulate matters smaller than 2.5 microns were analyzed using clustering techniques on a scale of days, months, days of a week and hours of a day. The results of hierarchical analysis by Ward's integration method showed that 365 days of the year can be divided into three separate clusters. Also 12 months of the year can be placed in 3 clusters.Months Azar, Aban and Shahrivar are in one cluster, and spring months and Mehr are in the other cluster. Tir and Mordad and months of winter are in the third cluster.Days of the week can be grouped into four categories: Friday and Sunday, respectively, the first and second clusters, and the third cluster, including Tuesdays and Thursdays, and finally, Saturdays, Mondays and Wednesdays. At hourly scale, the results of the study showed that the hours of daylight hours can be divided into homogeneous periods of dawn, morning, evening, and night. The lowest air pollution in the city of Mashhad is in the evening cluster (13-17h). Most of Mashhad air pollution is recorded in the night time, between 18 and 24 hours.