نشریه علوم و فنون نقشه برداری
سال سیزدهم شماره 1 (پیاپی 49، تابستان 1402)

Precipitable water vapor (PWV) is one of the most important data in meteorological studies. This component has high spatial and temporal changes, today the use of global navigation satellite systems (GNSS) observations is one of the ways to improve the accuracy of water vapor parameter estimation. The waves sent from GNSS satellites are delayed when passing through the troposphere layer. The troposphere delay is divided into two parts, dry and wet, and the wet part depends on changes in water vapor. In this article, the interpolation methods based on radial basis functions with 3D thin plate spline kernel, artificial neural network of perceptron type, kriging and inverse distance weighted have been evaluated. A region located in North America including 26 Global Positioning System (GPS) stations has been studied and the amounts of precipitable water vapor on two days in winter and summer have been evaluated using these data in the aforementioned methods. The value of the root mean square error (RMSE) using the 3D thin plate spline method for two days in winter and summer has been obtained as 0.6 and 1.62 mm, respectively, which has the lowest RMSE value compared to other methods. And as a result, there is a higher accuracy in both days. Finally, by using 3D thin plate spline interpolation method, a dense map of water vapor changes in the troposphere layer in the study area has been prepared, which can have an impact on forecasting the weather and estimating the amount of precipitation

Building extraction is one of the most crucial requirements of urban planning. Due to their availability and affordable cast, high-resolution remotely sensed images are often used for building extraction. Owing to their impressive performances, Deep learning techniques have attracted the attention of researchers for building extraction from high-resolution images. Nevertheless, most existing models perform poorly in recovering spatial details and discriminating buildings with various sizes and shapes. Hence, this paper proposes an improvement module to address the problems associated with multi-scale building extraction. The proposed module uses dilated convolutions to increase the receiving information area to reduce the discontinuities in the results of large buildings. Extracting large buildings using the proposed module and small buildings using the main architecture of the network has turned the proposed network into an effective method for building extraction. The results of the experiments showed that the proposed module with the IoU of 0.6495 and 0.8572 for Massachusetts and WHU data sets outperformed FCN, U-Net, USSP, and DeepLab V3+. The performance analysis of the proposed module also showed that this module was able to improve the performance of building extraction considering the IoU metric by 0.1077.

Motorcycle riders are one of the most dangerous transport users in the occurrence of crashes. According to economic conditions and traffic problems people are persuasion to use motorcycle for intra city transportation. The major purpose of this research is to present a method based on temporal analysis and cluster functions of the Geographical Information System (GIS) in order to analyze the statistical, spatial and temporal aspects of motorcycle crashes in order to identify roadway crash prone sections and provide necessary safety solutions. For this purpose, using Comap  and Geospatial functions, spatio-temporal patterns of motorcycle crashes and effect of the hourly and seasonal time on the location of crashes has been  investigated. According to the results of the research, it was found that the highest number of motorcycle crashes occur between 12:00 and 4:00 pm. In addition, The results showed that the highest number of motorcycle crashed happened in the fall and Also, the highest density of accidents is related to segments of Takhti Street. The results showed that using spatial cluster and temporal analysis is useful for identification of r roadway crash prone sections as well as the determination of the effective seasonal and hourly effects on crash clusters to perform adequate safety solutions related to motorcycles and apply preventively decisions based on the location and time of crashes.

The Ghotor doublet earthquake happened on 23 Feb 2020 near Khoy and Salmas cities in west Azerbaijan province of Iran near Iran-Turkey border .The first large event with a magnitude of 5.8 Mww (USGS) happened at 5:52 UTC (9:23 AM local time) and followed by a second large event of magnitude 6.0 Mw (USGS) at 16:00 UTC. The second event inflicted most of the building damages. No surface rupture has been reported for the events. We estimated the areal extent of the surface displacement related to the 2020 Ghotor doublet earthquakes using three sets of C-band imagery from the European Space Agency Sentinel 1A and 1B satellites. Due to small ground displacement, we could not model the fault geometry of the first mainshock. The ascending and descending displacement maps of the second main shock are used to jointly invert the causative fault plane parameters. To obtain the source parameters, we first down-sampled the unwrapped LOS surface displacements by a quadtree algorithm and then inverted the unwrapped interferograms to infer the geometry of a single rectangular plane with uniform slip in a uniform elastic half-space. The fault geometry parameters are location (X, Y, and Depth), size (length and width), orientation (strike, dip, and rake), and uniform slip of the rupture plane. We assume the X, Y, and depth to correspond to the center of the top edge of the rupture plane. We used a nonlinear inversion method as implemented in the open-source software called Geodetic Bayesian Inversion Software (GBIS) released by Centre for Observation and Modeling of Earthquakes, Volcanoes, and Tectonics (COMET). We used Okada‘s (1985) displacement green functions to model the displacement field. The calculated optimal model illustrates a northeast-striking (N24°) left-lateral rupture plane dipping ~86° towards the west. Once the geometry of the fault plane with uniform slip was estimated, we expanded the rupture plane 20 km along-strike and 12 km along down-dip directions and divided it into 1291 individual patches to obtain the distributed slip on the rupture plane. Each patch has a fixed geometry according to optimal source parameters obtained from the nonlinear modeling, and the slip was allowed to vary freely on the fault plane. We used a modified version of the open-source software called FaultResampler 1.4 to apply the linear inversion for calculating slip distribution on the rupture plane. The coseismic rupture concentrates around a center depth of 3 km with a maximum slip of 97±8  cm. Assuming a rigidity modulus of 30 GPa, the geodetic moment is estimated to be 1.517E+18 Nm, equivalent to a moment magnitude of 6.05 Mw.

Identification of changes in remote sensing images plays an important role in many applications, such as monitoring the growth of urbanization, land use changes, and assessing disasters and natural damages. This process aims to assign the label "changed" or "not changed" to the pixels of two images taken from the same place but at two different times. On the other hand, in the last decade, deep learning methods have attracted the attention of many researchers in this field due to their proper performance in interpreting and processing remote sensing data and the ability to remove feature engineering and extract high-level features from images. In this regard, in this article, an optimal deep learning model has been designed, which increases the accuracy of identifying changes in two-time images due to its hierarchical structure, appropriate efficiency of multiscale features, and effective design of feature transfer. Due to the optimal structure and architecture, the proposed model has higher speed and accuracy of results compared to some popular models such as BIT. Applying the proposed model to the two data sets under investigation indicates an average accuracy of 96% and less complex calculations.

With the increasing growth of positioning technologies and the use of navigation systems, a large volume of moving point object data, such as people, cars, ships, and animals, is available. However, the lack of integrity and incompleteness of these data for systemic, human, and environmental reasons challenges the analysis of trajectories and their effective application in various fields. Therefore, the reconstruction of missing data plays an important role in maximizing the capacity of movement data, particularly in navigation and track tracking. In this study, using the similarity measurement of trajectories approach, trajectories containing gaps are reconstructed. In this regard, the context-based dynamic time warping (CDTW) method, along with speed and direction movement parameters, are used to measure the similarity and reconstruct the trajectories of vessels in two regions of the Atlantic and Pacific Oceans. Two mechanisms, a constant number of trajectories and a specified threshold, are considered for reconstruction. The results show that using a constant number of trajectories in comparison with the specified threshold reduces the root mean square error (RMSE) and mean absolute error (MAE) from 1.5 and 1.4 to 0.5 and 0.4, respectively. In addition, increasing the length of the trajectories improves the RMSE and MAE values from 0.5 to 0.1 in the case of a constant number of trajectories and 1.5 to 0.3 in the case of the specified threshold.

Detecting small objects such as vehicles and swimming pools in high-spatial-resolution drone images is challenging due to their similar geometric and color features. The increase in the number of vehicles is not only a major challenge from the perspective of urban traffic but also leads to environmental problems such as pollution and warming. Therefore, monitoring these targets can play an important role in managing these problems. On the other hand, the construction and maintenance of swimming pools also require a significant amount of water, and monitoring these targets in urban environments is essential for water conservation. In this regard, drone remote sensing images and deep learning networks, which have a high ability to detect objects from these images, are considered suitable tools for monitoring these targets. Although valuable research has been done in this area to address each of the environmental challenges mentioned, there are still shortcomings in them. In this study, a new deep learning network YOLOv5+ has been developed to simultaneously detect two targets, vehicles and swimming pools, from drone images, in which the network's performance in extracting efficient features has been enhanced due to the use of the Inception mechanism in the intermediate layers. Additionally, in this study, DJI Mavic and DJI Mini Se drone data from Tianjin regions in China and the city of Cannes in France were used to evaluate the performance of the proposed network and compare it with the YOLOv5 and YOLOv7 deep learning networks. Finally, the results showed that the proposed network achieved an overall accuracy of 95% on the test set, which is an improvement of 2% over the YOLOv5 and YOLOv7 networks, indicating the efficiency of the approach proposed in this study.