Wildfires are destructive natural disasters occurring with ever increasing frequency and intensity with profound direct and indirect consequences. While direct impacts on life, assets damages, and losses are clear, indirect impacts, such as the health issues associated with prolonged smoke exposure, are more difficult to assess and thus mitigate in case of disaster. Earth Observation data can provide valuable insights into fire intensity on a large scale and in a timely manner. When combined with existing demographic information, we can gauge the potential health risk associated with wildfires. To automatically estimate the potential health risk during wildfires, we propose FireSENS, an algorithm which can aid warning and decision-making systems during wildfire events, thus enabling a more efficient and targeted approach to mitigate the health consequences.

Floods are the most impactful type of natural disaster with an ever increasing frequency and people at risk. Earth Observation data can help detect flood extents on a large scale in a timely manner. In this study we implement a Machine Learning algorithm consisting of a SENet and UNet to detect water and flood related damage in optical satellite data. The approach is applied to the devastating Pakistan floods from summer 2022 for which we trained three models and analysed the feasibility and transferability of the proposed approach. A locally trained model achieves excellent performance of IoU = 93.5% (Intersection over Union) while the best transferable model achieves IoU = 83.8%.

Flooding is a major hydro-meteorological event that impacts billions of people across the world daily. Models and Earth observation data are used for forecasting flood severity, extent, and depth, but these models and derived products are often not globally operational, and they often provide different outputs. Looking at recent disastrous events at a global level and the importance recently attributed to the need for early warning systems, it may seem that not much is being done to warn the public early enough, respond appropriately, or mitigate impacts. This is far from the truth. Worldwide, many organizations monitor hydrometeorological, geological and other types of hazards in order to allow the design, preparation and execution of an adequate response. The Model of Models (MoM) approach that leverages hydrologic models and Earth Observation datasets in an integrated manner is designed to provide flood risk information to assist emergency responders.

Achieving an efficient management of agricultural fields is crucial for farmers, considering the challenges posed by water resources sustainability. Monitoring tools allow winemakers to keep their vineyards under control and improve the plants' health with targeted actions, such as irrigations scheduling or specific treatments. Knowing the actual number of plants may not be evident, particularly in old plantation, where vines might have been removed or added to the initial planting scheme, or even they have never been counted. However, the knowledge of the number and position of the plants is important to be able to adapt the irrigation network and properly adjust the irrigation schedule. The aim of this project is to build a drone service for precision monitoring of vineyards. Here, an algorithm to detect the position and number of plants in vineyards using drone RGB imagery is presented. First results show a plant detection accuracy of 87%.

With the ongoing proliferation of satellite data, in particular open-access satellite imagery, from both optical and synthetic aperture radar (SAR) sensors, the number of downstream applications is rapidly growing. Developers of Earth Observation (EO)-based products and services, as well as expert and non-expert users of such tools, thus need access to a cloud computing infrastructure offering interoperable analysis functionality. Here, we present the versatility of such a cloud-based infrastructure called WASDI. WASDI, a web-advanced space development interface, is an online EO analytics platform where EO experts can develop and deploy applications (apps) and users can use them to processes satellite images on demand to generate value-added content.

Flooding is the most common natural hazard worldwide, affecting over a billion people and costing $100 billion every year. This will likely increase in the future due to increasing population and assets in the flood-prone zones and climate change. Earth observation data have been utilized to map flooding on a global scale. This was mostly done by utilizing optical bands, as for these satellites the return period is relatively short; often daily. However, the usage of optical data can be restrictive due to e.g. cloud cover and nighttime. With synthetic aperture radar data most of these limitations can be overcome. As such, many entities now provide remotely sensed flood products, such that it becomes difficult for users to determine the best available flood data source. Here we describe the long-time development and implementation of a semi operational `one-stop-shop' portal that contains globally scoped, flood prediction, monitoring capabilities and risk evaluations by leveraging on efforts of the entire flood community.

In our changing world, floods are a threat of increasing concern. Within this context, flood mapping is important for both damage assessment and forecast improvement. Due to the suitability of synthetic aperture radar (SAR) for flood mapping, a broad range of SAR-based flood mapping algorithms has been developed during the past years. However, most of these algorithms were presented based on a single test case only and comparisons between methods are rare. This paper presents an in-depth assessment and comparison of the established pixel-based flood mapping approaches, including global and enhanced thresholding, active contour modeling and change detection. The methods were tested on medium-resolution SAR images of different flood events and lakes across the U.K. and Ireland and were evaluated on both accuracy and robustness. Results indicate that the most suited method depends on the area of interest and its characteristics as well as the intended use of the observation product. Due to its high robustness and good performance, tiled thresholding is suited for automated, near-real time flood detection and monitoring. Active contour models can provide higher accuracies but require long computation times that strongly increase with increasing image sizes, making them more appropriate for accurate flood mapping in smaller areas of interest.

We demonstrate the complementarity of a multitude of satellite flood maps and large-scale flood inundation modeling. We employ a unique set of maps, from both optical and radar imagery, that were delivered to emergency responders during the Texas flood disaster of late May, early June 2015. Specifically, for this study, a two-dimensional hydrodynamic model was built to simulate the best possible inundation re-analysis of the flood event in locations along the major rivers, including urban and coastal settings. Subsequently, integrating the model event re-analysis and the satellite flood data demonstrated the unique complementarity of these available multi-temporal and multi-resolution imagery and the large-scale inundation model. This allowed a thorough assessment of the uncertainty and value of “big” Earth Observation data for flood disaster response and for integration with flood modeling for effective event re-analysis, which we anticipate can help guide better flood resilience planning.

Very high resolution synthetic aperture radar (SAR) sensors represent an alternative to aerial photography for delineating floods in built-up environments where flood risk is highest. However, even with currently available SAR image resolutions of 3 m and higher, signal returns from man-made structures hamper the accurate mapping of flooded areas. Enhanced image processing algorithms and a better exploitation of image archives are required to facilitate the use of microwave remote-sensing data for monitoring flood dynamics in urban areas. In this paper, a hybrid methodology combining backscatter thresholding, region growing, and change detection (CD) is introduced as an approach enabling the automated, objective, and reliable flood extent extraction from very high resolution urban SAR images. The method is based on the calibration of a statistical distribution of “open water” backscatter values from images of floods. Images acquired during dry conditions enable the identification of areas that are not “visible” to the sensor (i.e., regions affected by “shadow”) and that systematically behave as specular reflectors (e.g., smooth tarmac, permanent water bodies). CD with respect to a reference image thereby reduces overdetection of inundated areas. A case study of the July 2007 Severn River flood (UK) observed by airborne photography and the very high resolution SAR sensor on board TerraSAR-X highlights advantages and limitations of the method. Even though the proposed fully automated SAR-based flood-mapping technique overcomes some limitations of previous methods, further technological and methodological improvements are necessary for SAR-based flood detection in urban areas to match the mapping capability of high-quality aerial photography.

A near real-time flood detection algorithm giving a synoptic overview of the extent of flooding in both urban and rural areas, and capable of working during night-time and day-time even if cloud was present, could be a useful tool for operational flood relief management. The paper describes an automatic algorithm using high-resolution synthetic aperture radar (SAR) satellite data that builds on existing approaches, including the use of image segmentation techniques prior to object classification to cope with the very large number of pixels in these scenes. Flood detection in urban areas is guided by the flood extent derived in adjacent rural areas. The algorithm assumes that high-resolution topographic height data are available for at least the urban areas of the scene, in order that a SAR simulator may be used to estimate areas of radar shadow and layover. The algorithm proved capable of detecting flooding in rural areas using TerraSAR-X with good accuracy, classifying 89% of flooded pixels correctly, with an associated false positive rate of 6%. Of the urban water pixels visible to TerraSAR-X, 75% were correctly detected, with a false positive rate of 24%. If all urban water pixels were considered, including those in shadow and layover regions, these figures fell to 57% and 18%, respectively.

Flooding is a major hazard in both rural and urban areas worldwide, but it is in urban areas that the impacts are most severe. An investigation of the ability of high-resolution TerraSAR-X synthetic aperture radar (SAR) data to detect flooded regions in urban areas is described. The study uses a TerraSAR-X image of a one-in-150-year flood near Tewkesbury, U.K., in 2007, for which contemporaneous aerial photography exists for validation. The German Aerospace Center (DLR) SAR end-to-end simulator (SETES) was used in conjunction with airborne scanning laser altimetry (LiDAR) data to estimate regions of the image in which water would not be visible due to shadow or layover caused by buildings and taller vegetation. A semiautomatic algorithm for the detection of floodwater in urban areas is described, together with its validation using aerial photographs. Of the urban water pixels that are visible to TerraSAR-X, 76% were correctly detected, with an associated false positive rate of 25%. If all the urban water pixels were considered, including those in shadow and layover regions, these figures fell to 58% and 19%, respectively. The algorithm is aimed at producing urban flood extents with which to calibrate and validate urban flood inundation models, and these findings indicate that TerraSAR-X is capable of providing useful data for this purpose.

Exploitation of river inundation satellite images, particularly for operational applications, is mostly restricted to flood extent mapping. However, there lies significant potential for improvement in a 3-D characterization of floods (i.e., flood depth maps) and an integration of the remote-sensing-derived (RSD) characteristics in hydraulic models. This paper aims at developing synthetic aperture radar (SAR) image analysis methods that go beyond flood extent mapping to assess the potential of these images in the spatiotemporal characterization of flood events. To meet this aim, two research issues were addressed. The first issue relates to water level estimation. The proposed method, which is an adaptation to SAR images of the method developed for water level estimation using flood aerial photographs, is composed of three steps: (1) extraction of flood extent limits that are relevant for water level estimation; (2) water level estimation by merging relevant limits with a Digital Elevation Model; and (3) constraining of the water level estimates using hydraulic coherence concepts. Applied to an ENVISAT image of an Alzette River flood (2003, Grand Duchy of Luxembourg), this provides plusmn54-cm average vertical uncertainty water levels that were validated using a sample of ground surveyed high water marks. The second issue aims at better constraining hydraulic models using these RSD water levels. To meet this aim, a "traditional" calibration using recorded hydrographs is completed via comparison between simulated and RSD water levels. This integration of the RSD characteristics proves to better constrain the model (i.e., the number of parameter sets providing acceptable results with respect to observations has been reduced). Furthermore, simulations of a flood event of a different return period (2007) using the model calibrated for the 2003 flood event shows the reliability of the latter for flood forecasting.

On December 12, 2006, both the European Remote Sensing Satellite 2 and Environmental Satellite recorded a high-magnitude flood event on the River Dee in Wales (U.K.) only 28 min apart. This unique opportunity enables the creation of a very rare but extremely useful observed data set for flood inundation studies. For flood management purposes, hydrodynamic models are often run after an event but with field data gauged during the event to approximate both flood area and depth. As an adequateaprioridefinition of model parameters is difficult, they tend to be run with multiple parameter sets to generate a likelihood of inundation map. However, as field observations of events are often very scarce, these output maps cannot be validated with field-observed probabilities. This paper illustrates how this unique set of spaceborne radar images can be used in combination with five widely used image processing techniques to generate an event-specific inundation map that expresses a degree of belief that a given pixel is possibly flooded. It is expected that the value of this multialgorithm ensemble-based map opens up new ways to evaluate the performance of hydrodynamic models, as it contains information which has, to the authors' knowledge, not previously been available.

The present case study focuses on monitoring the wetness state of the experimental Bibeschbach catchment (10.8 km2), located within the Alzette river basin in the Grand-Duchy of Luxemburg over the last three winters (2005-2008). The objectives of this study are (1) to retrieve soil moisture from spaceborne active and passive microwave sensors, namely AMSR-E and ERS-2 SAR, (2) to compare the remote sensing-derived estimates of basin-averaged soil moisture with ground measurements that are performed throughout the catchment.

Observed spatially distributed water stages with uncertainty are of considerable importance for flood modeling and management purposes but are difficult to collect in the field during a flood event. Synthetic aperture radar (SAR) remote sensing offers an inviting alternative to provide this kind of data. A straightforward technique to derive water stages from a single SAR flood image is to extract heights from a digital elevation model at the flood boundaries. Schumann et al. have presented a regression modeling approach as an improvement to this simple technique. However, regression modeling associated with their model may restrict output to mapping purposes rather than extend it to integration with other data or models. This letter introduces an inviting alternative that conducts statistical analysis on river cross-sectional data points, thereby allowing uncertainty assessment of remote-sensing-derived water stages without any regression modeling constraint. This renders remote-sensing data fit for, e.g., flood inundation model evaluation with uncertainty in observations and data assimilation studies, where (linear) ldquotransformation,rdquo i.e., modeling, to observed data should be minimal.

This paper presents a remote-sensing-based steady-state flood inundation model to improve preventive flood-management strategies and flood disaster management. The Regression and Elevation-based Flood Information eXtraction (REFIX) model is based on regression analysis and uses a remotely sensed flood extent and a high-resolution floodplain digital elevation model to compute flood depths for a given flood event. The root mean squared error of the REFIX, compared to ground-surveyed high water marks, is 18 cm for the January 2003 flood event on the River Alzette floodplain (G.D. of Luxembourg), on which the model is developed. Applying the same methodology on a reach of the River Mosel, France, shows that for some more complex river configurations (in this case, a meandering river reach that contains a number of hydraulic structures), piecewise regression is required to yield more accurate flood water-line estimations. A comparison with a simulation from the Hydrologic Engineering Centers River Analysis System hydraulic flood model, calibrated on the same events, shows that, for both events, the REFIX model approximates the water line reliably