Revista de Teledetección

Quantitative analysis of subsidence in the Southwestern Sabana de Bogotá using InSAR geodesy

Fredy Diaz-Mila — 2024-10-21

Subsidence is a geological phenomenon that consists of the gradual sinking of the earth’s surface, which can be caused by natural actions or by human activity. The Satellite Applications Group for the Study of Earth Dynamics (ASEDT) of the Geohazards Directorate of the Colombian Geological Survey, under the framework of the GeoRED project (Geodesy: Deformation Studies Network), using InSAR (Interferometric Synthetic Aperture Radar) geodesy techniques, has quantified the phenomenon of subsidence in 13 municipalities that are part of the Sabana de Bogotá, using Sentinel-1 interferometric images (184 of descending orbit and 225 of ascending orbit), for the period from October 2014 to December 2021. A total of 840 interferograms were generated, of which 345 correspond to ascending orbits and 495 to descending orbits, which allowed estimating the values of the movements along the line-of-sight (LOS) for each set of images, and subsequently using combination techniques the vertical and horizontal east-west velocities were estimated. One of the municipalities with the greatest subsidence in the study area is El Rosal, where a vertical displacement rate of up to 12 cm/year is estimated, i.e. a value approximately four times higher than the estimate made by Mora-Páez et al. (2021) in the city of Bogotá.

Assessment of the impact of Hurricane Lorena (2019) on the mangrove forest of Espiritu Santo archipelago, Mexico using Sentinel-2 imagery

Daniel A. Robles-Archundia — 2024-10-28

Mangrove ecosystems are a priority for conservation. They provide diverse ecosystem services and are key in the life cycle of many species. However, they are threatened by various productive activities and natural phenomena such as hurricanes, which impact the coasts causing damage to the vegetation. Monitoring the effects of hurricane impact on mangroves is a complex task since many resources are needed to access the devastated sites and conduct evaluations over large areas. Therefore, remote sensing products represent tools with great potential to assess the most vulnerable areas. In this study, the impact of Hurricane Lorena which hit the archipelago of Espiritu Santo, BCS, in the summer of 2019 was evaluated. Two Sentinel-2 satellite images taken before (09/09/2019) and after (09/24/2019) the hurricane passed were used. Four vegetation indices (VI) related to photosynthetic activity and canopy moisture content were calculated. In addition, ΔVI was calculated for each index, which represents the proportional reduction of the VI value after the impact. The results showed a general increase in the values of the four VI throughout the study area. The category of hurricane Lorena, which confers a lower wind speed and the precipitation associated with this meteor, could explain the increase in the values of the VI. The differentiated response among the four VI shows the importance of using more than one indicator in studies assessing the impact of natural phenomena on coastal vegetation.

Convolutional neural network-based semantic segmentation model for land cover classification in páramo ecosystems

Marcela Reyes — 2025-01-15

Páramo ecosystems are essential for water regulation and biodiversity conservation in mountainous areas. However, they face significant threats due to climate change and human activities such as agriculture, livestock farming, and mining. The absence of clear boundaries and continuous monitoring systems for their land cover hinders effective protection highlighting the need to employ advanced digital techniques that provide highly accurate, up-to-date information. Convolutional neural networks (CNNs) have emerged as promising tools for semantic segmentation of satellite images. This research aimed to evaluate the performance of two CNNs architectures U-Net++ and DeepLabV3+ for land cover classification in the Tota-Bijagual-Mamapacha (TBM) páramo complex in Colombia, using Landsat 8 imagery from 2017 to 2019 and land cover labels from 1:100.000, national coverage map produced by IDEAM in 2018. The results showed U-Net++ achieved a kappa of 0.60, while DeepLabV3+ obtained a kappa of 0.59. In páramo covers, U-Net++ achieved an F1 of 78.43% for Herbazal and 79.22% for Forests, while DeepLabV3+ achieved F1 of 75% and 74.27%, respectively, confirming the potential of CNNs for land cover classification in these ecosystems. Although both models presented similar processing times, class imbalance and reliance on consistent labels affected their performance in heterogeneous covers. This research establishes a methodological foundation for future studies and suggests addressing these limitations to
improve efficiency and thematic accuracy in the classification and monitoring in páramo ecosystems.

Using satellite imagery to assess the glacier retreat in King George Island, Antarctica

Ibeth Rojas-Macedo — 2025-01-15

In recent decades, remote sensing has become a powerful tool for continuously monitoring glacier dynamics in remote areas, enabling the identification of significant spatiotemporal changes due to its capacity to provide multitemporal information at regional and global scales. In this study, Landsat satellite images (1989–2020) were used to quantify glacier retreat in the ice cap of King George Island (KGI), located in the Antarctic Peninsula, and to evaluate the teleconnections of El Niño – Southern Oscillation - ENSO (ONI and SOI indices) with climatic
variables (temperature and precipitation) in this region. Our findings reveal a 10% loss in glacier coverage over the last 31 years, with a slower glacier retreat observed since 2008. Glaciers with smaller areas and marine terminating were the most affected. Of the 73 glaciers on KGI, 42% had continental terminating, 21% had marine terminating, and 37% had mixed terminating (continental and marine). Of the total glacier area lost, 35% corresponds to glaciers with marine terminating, while 16% corresponds to glaciers with continental terminating. Furthermore, climatic variables exhibited heterogeneous responses during ENSO events, with a significant correlation between mean temperature and ONI at the annual level and during the austral spring, which may be influencing glacier retreat in the study area to some extent.

Geospatial flood prospecting in a capital city on the Gulf of Mexico

Andrés De la Rosa Portilla — 2025-01-07

Environmental risks are a concern for the habitability of the planet. It is estimated that by 2050, 68 % of the world’s population will live in cities, and in Mexico, 79 % already reside in urban areas. This population concentration faces threats such as climate variability and the effects of climate change, which manifested themselves in extreme weather events, including flooding. The Topographic Wetness Index (TWI) is a useful tool for identifying areas susceptible to flooding and planning preventive infrastructure. In this research, a diachronic digital hemerographic analysis of flooding in Xalapa de Enriquez, Veracruz, Mexico; was performed by modeling the TWI with LIDAR data to identify flood-prone areas. Marginalization, per capita tree cover and storm drainage in the city were also analyzed. The results showed that, in 18 years, 369 floods occurred in 120 neighborhoods, affecting 61.5 % of the inhabitants. 56 % of the floods occurred between 2017 and 2022. Two of the three colonias with the most floods have a medium degree of marginalization. 39.8 % of the colonias have less than 5 m² of trees per inhabitant, failing to comply with local legislation, and 56.7 % do not reach the internationally recommended 15 m². The TWI revealed that 45 % of the floodable surface is not registered in the Flood Hazards and Vulnerability section of the Municipal Land Use Planning Program. The 10 neighborhoods with the most flooded have the lowest percentages of storm sewers. It is concluded that it is feasible and desirable to replicate the TWI in other cities to prevent flooding and plan integrated disaster risk management works, valuing the per capita tree-covered area in coordination with improvements in storm drainage, through an intra-urban neighborhood approach.

Optimizing SVM for argan tree classification using Sentinel-2 data: A case study in the Sous-Massa Region, Morocco

Abdelhak El Kharki — 2024-11-04

The development of efficient classifiers for land cover remains challenging due to the presence of hyperparameters in the model. Conventional approaches rely on manual tuning, which is both time-consuming and impractical, often leading to suboptimal results. This study aimed to optimize the hyperparameters of the Support Vector Machine (SVM) algorithm using the grid search method to map the distribution of the Argan forest in the Souss-Massa region of Morocco from Sentinel-2 satellite image. To achieve this, we examined the C parameter for the linear function, as well as the C and gamma parameters for the radial RBF and sigmoid functions. Similarly, we explored the C, gamma, and degree parameters for the polynomial function chosen using the grid search method. These parameters are compared with the default hyperparameters of each SVM function. The results are validated using the cross-validation method and by the following scores: accuracy, precision, recall, F1 score, and Cohen’s Kappa. The experiments were conducted using the Earth Engine Python API in Google Colab (Google Collaboratory). In addition, experimental results indicate that the hyperparameters selected by grid search yield higher scores than the default hyperparameters. The best results were achieved using the hyperparameters of the polynomial base kernel, specifically with C = 10, degree = 2, and gamma = 10. Accuracy = 96.61%.