Evaluación de algoritmos de clasificación en la plataforma Google Earth Engine para la identificación y detección de cambios de construcciones rurales y periurbanas a partir de imágenes de alta resolución

<p>Building change detection based on remote sensing imagery is a key task for land management and planning e.g., detection of illegal settlements, updating land records and disaster response. Under the post- classification comparison approach, this research aimed to evaluate the feasibility of several classification algorithms to identify and capture buildings and their change between two time steps using very-high resolution images (&lt;1 m/pixel) across rural areas and urban/rural perimeter boundaries. Through an App implemented on the Google Earth Engine (GEE) platform, we selected two study areas in Colombia with different images and input data. In total, eight traditional classification algorithms, three unsupervised (K-means, X-Means y Cascade K-Means) and five supervised (Random Forest, Support Vector Machine, Naive Bayes, GMO maximum Entropy and Minimum distance) available at GEE were trained. Additionally, a deep neural network named Feature Pyramid Networks (FPN) was added and trained using a pre-trained model, EfficientNetB3 model. Three evaluation zones per study area were proposed to quantify the performance of the algorithms through the Intersection over Union (IoU) metric. This metric, with a range between 0 and 1, represents the degree of overlapping between two regions, where the higher agreement the higher IoU values. The results indicate that the models configured with the FPN network have the best performance followed by the traditional supervised algorithms. The performance differences were specific to the study area. For the rural area, the best FPN configuration obtained an IoU averaged for both time steps of 0.4, being this four times higher than the best supervised model, Support Vector Machines using a linear kernel with an average IoU of 0.1. Regarding the setting of urban/rural perimeter boundaries, this difference was less marked, having an average IoU of 0.53 in comparison to 0.38 obtained by the best supervised classification model, in this case Random Forest. The results are relevant for institutions tracking the dynamics of building areas from cloud computing platfo future assessments of classifiers in likewise platforms in other contexts.</p>

Mapping Sandy Areas and their changes using remote sensing. A Case Study at North-East Al-Muthanna Province, South of Iraq

<p>Sandy areas are the main problem in regions of arid and semi-arid climate in the world that threaten urban life, buildings, agricultural, and even human health. Remote sensing is one of the technologies that can be used as an effective tool in dynamic features study of sandy areas and sand accumulations. In this study, two new indices were developed to separate the sandy areas from the non-sandy areas. The first one is called the Normalized Differential Sandy Areas Index (NDSAI) that has been based on the assumption that the sandy area has the lowest water content (moisture) than the other land cover classes. The second other is called the Sandy Areas Surface Temperature index (SASTI) which was built on the assumption that the surface temperature of sandy soil is the highest. The results of proposed indices have been compared with two indices that were previously proposed by other researchers, namely the Normalized Differential Sand Dune Index NDSI and the Eolain Mapping Index (EMI). The accuracy assessment of the sandy indices showed that the NDSAI provides very good performance with an overall accuracy of 89 %. The SASTI can isolate many sandy and non-sandy pixels with an overall accuracy about 86 %. The performance of the NDSI is low with an overall accuracy about 82 %. It fails to classify or isolate the vegetation area from the sandy area and might have better performance in desert environments. The performing of NDSAI that is calculated with the SWIR1 band of the Landsat satellite is better than the performing of NDSI that is calculated with the SWIR2 band of the same satellite. EMI performance is less robust than other methods as it is not useful for extracting sandy surfaces in area with different land covers. Change detection techniques were used by comparing the areas of the sandy lands for the periods from 1987 to 2017. The results showed an increase in sandy areas over four decades. The percentage of this increase was about 20 % to 30 % during 2002 and 2017 compared to 1987.</p>

Aplicación de una metodología simplificada para estimar las curvas de capacidad de embalses

<p>In the management of the regulation volumes of the water resource projected for a variety of benefits (hydroelectric uses, irrigation, drinking water, among others), it is essential to calculate the morphometric variables of the reservoirs to anticipate changes in their morphology and predict how these changes could affect projected achievement. In this document, taking into consideration the fundamental concepts of Integral Calculus, the development of an innovative methodology is presented to obtain the Cota-Volume and Cota-Area curves in reservoirs; the methodology was formulated in R programming language with the help of geographic information tools. A computational optimization was achieved for the processing of the variables of level, area and volume of a regulation body respect to the use of traditional methodologies. To validate the developed tool, the capacity curves of regulation volume of the Minas - San Francisco reservoir located in the south of the Republic of Ecuador were obtained. This reservoir was designed to dislodge its sediments with washing processes. This fact has motivated the continuous monitoring of the morphological conditions of the reservoir to plan maintenance processes due to the loss of volume and to the deposit of particles from its tributaries. In addition, an analysis based on wavelets curves was applied to the digital elevation models obtained from LiDAR techniques and bathymetric echo sounder to demonstrate the sedimentation processes that occur in this body of regulation.</p>

Monitoreo de la variación del almacenamiento de agua en la cuenca del Medio y Bajo Paraná a partir de datos GRACE, GRACE FO, TRMM y GLDAS

<p>GRACE (Gravity Recovery and Climate Experiment) is a satellite mission that can monitor mass distributions in the Earth system, which is closely related to the consequences of climate change. This gravimetric satellite allows to obtain monthly variations of the Earth’s gravity field, which can be associated with water mass variations, after removing the effects of oceanic tides and solid Earth, as well as non-tidal oceanic and atmospheric contributions. In this work, data from GRACE (2002-2017) and GRACE FO (since 2018) were used to analyze the variation of the water mass in the Middle and Low Paraná river basin. The interpretation of the results was carried out by associating the mass anomalies derived from GRACE data with information from the TRMM global rainfall mission. Monthly maps of GRACE water mass variations and TRMM precipitation were produced, which made possible a thorough analysis at a regional level of this mass redistribution in the basin, and its connection to the El Niño and La Niña events that took place in the period under study. The water deficits shown in the 2009 GRACE maps are, in fact, related to the intense episode of La Niña that occurred in the period 2008-2009; while the excess of water storage depicted on the 2016 and 2019 maps is connected to the El Niño phenomenon. Moreover, GRACE has also detected drought events in different sectors between 2011-2012, together with floods in the years 2007 and 2010. Monthly GRACE-derived water storage changes were compared with the independent components of the water balance in the region using different hydrological models estimates. Finally, the temporal variations of the groundwater and the soil part (surface water, soil moisture) were analyzed using the Global Land Data Assimilation System GLDAS. The variables showed a good correlation between them, reaching values of <span> ~</span>r = 0.80.</p>

Estimación de la evapotranspiración del cultivo de arroz en Perú mediante el algoritmo METRIC e imágenes VANT

Modern remote measurement techniques using cameras mounted on an unmanned aerial vehicle (UAV) have made possible to acquire high-resolution images and estimating evapotranspiration at more detailed spatial and temporal scales. The objective of the present research was to estimate crop evapotranspiration (ETc) of rice crop using the “mapping evapotranspiration with internalized calibration model (METRIC)” using high spatial resolution multispectral and thermal images obtained from a UAV. A total of 18 flights with UAV were performed to get the images; likewise, data were collected from the weather station and thermocouple information installed in the crop canopy under soil water potential conditions of –10 kPa (T1), –15 kPa (T2), –20 kPa (T3) and a control of 0 kPa (T0), from November 13, 2017, to April 30, 2018. The results indicate that the METRIC model compared to ETc measurements recorded by a field drainage lysimeter presents a Pearson correlation coefficient (r) of 0.97, root mean square error (RMSE) of 0.51 mm d^–1, Nash-Sutcliffe coefficient (EF) of 0.87 and underestimation of 7 %. Evapotranspiration reached values of 7.48 mm d^–1, with differences between treatments of 0.2 %, 6 % and 8 % concerning to T0 and yield reduction of 9 %, 34 % and 35 % for T1, T2 and T3 soil water potential. The high[1]resolution images allowed obtaining detailed information on the spatial variability of ETc that could be used in the more efficient application of plot irrigation.

Integración de imágenes de sensores remotos en el desarrollo de indicadores medioambientales en cuencas mediterráneas. Aplicación al seguimiento de su estado hídrico y productividad

<p>This thesis proposes the use of remote sensing images of different spatial, spectral and temporal resolutions that, combined with meteorological, hydrological and phenological data, can be used to produce indicators of different ecosystem variables related to productivity and water status in different unique systems of the Mediterranean region. Specifically, the development of three indicators closely linked to each other is proposed: an indicator of the water status of the soil at the end of the dry season from the state of different vegetation covers; an indicator of the productivity of natural pastures, the main food support for extensive livestock in dehesa ecosystems, based on their status and the climatic conditions of the period evaluated; and, finally, an indicator of the relationship between water state of the soil and the natural pasture phenological state.</p>

Surface water extent dynamics from three periods of continuous Landsat time series; subregional differences across Argentine plains

<p>The Pampean region in Argentina is an extensive plain characterized by abundant shallow lakes that fulfill many environmental, ecological, and social functions. This study aims to detect the multiannual lake area changes in this region during 2001-2009 using remote sensing, including lakes as small as ≥10,000 m² or 1 ha. Landsat scenes of the wet (2008-2009), normal (2006), and dry (2008-2009) seasons were obtained, and using remote sensing techniques, the number and area of shallow lakes were calculated. The spatiotemporal variation of shallow lakes was studied in different climate periods in eight singular subregions. Spatial associations between annual precipitation and lake number and area were analyzed through the development of a Geographic Information System (GIS) at a subregional scale. During the study period the total lake area in the Pampean region decreased by 5257.39 km²(62 %), but each subregion showed different responses to climatic events. In seven of them, the differences between climate periods prove to be statistically significant (P&gt;0.01). The relationship between precipitation and lake number and area revealed the domain of positive association. We conclude that climate factors play a dominant role in lake changes across the Pampean plains. However, other factors such as origin, topographic and edaphic characteristics intensify or mitigate changes in surface hydrology.</p>

Evaluación de parámetros de segmentación en OBIA para la clasificación de coberturas del suelo a partir de imágenes VANT

<p>Unmanned Aerial Vehicles (UAVs) have given a new boost to remote sensing and image classification techniques due to the high level of detail among other factors. Object-based image analysis (OBIA) could improve classification accuracy unlike to pixel-based, especially in high-resolution images. OBIA application for image classification consists of three stages i.e., segmentation, class definition and training polygons, and classification. However, defining the parameters: spatial radius (SR), range radius (RR) and minimum region size (MR) is necessary during the segmentation stage. Despite their relevance, they are usually visually adjusted, which leads to a subjective interpretation. Therefore, it is of utmost importance to generate knowledge focused on evaluating combinations of these parameters. This study describes the use of the mean-shift segmentation algorithm followed by <em>Random Forest </em>classifier using Orfeo Toolbox software. It was considered a multispectral orthomosaic derived from UAV to generate a suburban map land cover in town of El Pueblito, Durango, Mexico. The main aim was to evaluate efficiency and segmentation quality of nine parameter combinations previously reported in scientific studies.This in terms of number generated polygons, processing time, discrepancy measures for segmentation and classification accuracy metrics. Results evidenced the importance of calibrating the input parameters in the segmentation algorithms. Best combination was RE=5, RR=7 and TMR=250, with a Kappa index of 0.90 and shortest processing time. On the other hand, RR showed a strong and inversely proportional degree of association regarding the classification accuracy metrics.</p>

Determinación de la temperatura de la superficie terrestre mediante imágenes Landsat 8: Estudio comparativo de algoritmos sobre la ciudad de Granada

<p>The use of satellite images has become, in recent decades, one of the most common ways to determine the Land Surface Temperature (LST). One of them is through the use of Landsat 8 images that requires the use of single-channel (MC) and two-channel (BC) algorithms. In this study, the LST of a medium-sized city, Granada (Spain) has been determined over a year by using five Landsat 8 algorithms that are subsequently compared with ambient temperatures. Few studies compare the data source with the seasonal variations of the same metropolis, which together with its geographical location, high pollution and the significant thermal variations it experiences make it a suitable place for the development of this research. As a result of the statistical analysis process, the regression coefficients R², mean square error (RMSE), mean error bias (MBE) and standard deviation (SD) were obtained. The average results obtained reveal that the LST derived from the BC algorithms (1.0 °C) are the closest to the ambient temperatures in contrast to the MC (-5.6 °C), although important variations have been verified between the different zones of the city according to its coverage and seasonal periods. Therefore, it is concluded that the BC algorithms are the most suitable for recovering the LST of the city under study.</p>

Estudio de las fluctuaciones del nivel del agua en la laguna de Gallocanta (Aragón, España) mediante imágenes satelitales de Sentinel-2

<p>This study has been monitored for five years by Sentinel-2 satellite images, at different seasons of the year, of the fluctuations in the water level of the Gallocanta Lake (between the provinces of Teruel and Zaragoza, Spain) considered a hypersaline and endorheic wetland, which has characteristics that make it unique in the geographical area in which it is located, as well as for the operation of the system. Rainfall in the area has a wide variation giving the maximums in the moths of May and June and the minimums in January and February, with considerable fluctuations in the water level from the almost total drying of the lagoon to the filling with a depth of approximately 3 meters.</p>