Session GI6.1

[Programme]

GI6.1 | Remote sensing for environmental monitoring

Remote sensing for environmental monitoring

Convener: Annalisa CappelloECSECS | Co-conveners: Gaetana Ganci, Gabor Kereszturi, Veronika Kopackova

Orals

| Tue, 16 Apr, 08:30–10:15 (CEST)

Room -2.16

Posters on site

| Attendance Wed, 17 Apr, 16:15–18:00 (CEST) | Display Wed, 17 Apr, 14:00–18:00

Hall X4

Posters virtual

| Attendance Wed, 17 Apr, 14:00–15:45 (CEST) | Display Wed, 17 Apr, 08:30–18:00

vHall X4

Remote sensing measurements, acquired using different platforms - ground, UAV, aircraft and satellite - have increasingly become rapidly developing technologies to study and monitor Earth surface, to perform comprehensive analysis and modeling, with the final goal of supporting decision systems for ecosystem management. The spectral, spatial and temporal resolutions of remote sensors have been continuously improving, making environmental remote sensing more accurate and comprehensive than ever before. Such progress enables understanding multiscale aspects of high-risk natural phenomena and development of multi-platform and inter-disciplinary surveillance monitoring tools. The session welcomes contributions focusing on present and future perspectives in environmental remote sensing, from multispectral/hyperspectral optical and thermal sensors. Applications are encouraged to cover, but not limited to, the monitoring and characterization of environmental changes and natural hazards from volcanic and seismic processes, landslides, and soil science. Specifically, we are looking for novel solutions and approaches including the topics as follows: (i) state-of-the-art techniques focusing on novel quantitative methods; (ii) new applications for state-of-the-art sensors, including UAVs and other close-range systems; (iii) techniques for multiplatform data fusion.

Orals: Tue, 16 Apr | Room -2.16

Chairpersons: Annalisa Cappello, Gaetana Ganci, Veronika Kopackova

08:30–08:35

5-minute convener introduction

08:35–08:45

EGU24-10884

ECS

Highlight

On-site presentation

Hyperspectral Analysis for Protected Agriculture land cover mapping: A Remote Sensing Approach

Davide Parmeggiani, Francesca Despini, Sofia Costanzini, Sergio Teggi, and Daniele la Cecilia

Spaceborne and airborne remote sensing data serve as powerful tools for the analysis and
monitoring of both urban and agricultural territories, with diverse applications contingent upon spatial
resolution. In recent years, remote sensing imagery has been utilized for the recognition of protected
agriculture landcovers, such as greenhouses and mulch. Various studies in the scientific literature have
focused on satellite sensors like Sentinel-2 and WorldView-3, mapping the presence of protected
agriculture surfaces and implementing specific indices for recognition.
A recurrent limitation in these studies lies in the often insufficient spatial resolution of the sensors,
particularly for identifying smaller-sized greenhouses. Additionally, spectral resolution is crucial. While
some laboratory studies analyse the spectral characteristics of plastic surfaces typical of protected
agriculture, they often neglect the issue of mixed pixels inherent in satellite or aerial detection.
The aim of this study is to analyze images from the AVIRIS airborne sensor over the agricultural area of
Salerno in southern Italy. AVIRIS, a hyperspectral sensor with over 400 bands covering the visible (VIS) to
the shortwave infrared (SWIR) region, provided images with a spatial resolution of 1m and 3m. We
scrutinize these images to discern the spectral signatures of different types of greenhouses in the study
area, subsequently comparing them with other land cover classes. For this, we employ supportive tools,
including specific spectral indices and transformations such as Tasselled Cap and Principal Components
Analysis (PCA). We implement the Region of Interest (ROI) separability technique to identify distinctive
spectral features in the signatures of protected agriculture coverings that differentiate them from other
surfaces. Finally, the spectral signatures obtained from AVIRIS offer the opportunity to simulate spectral
responses of other satellite sensors with lower spatial and/or spectral resolutions, assessing the suitability
of currently available data for recognizing this specific type of surface.

How to cite: Parmeggiani, D., Despini, F., Costanzini, S., Teggi, S., and la Cecilia, D.: Hyperspectral Analysis for Protected Agriculture land cover mapping: A Remote Sensing Approach, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10884, https://doi.org/10.5194/egusphere-egu24-10884, 2024.

08:45–08:55

EGU24-12801

On-site presentation

Relative performance of super-resolved Sentinel-2 images for built-up area mapping using deep learning

Misganu Debella-Gilo

Early detection of built-up areas is important for planning and understanding the impacts of urbanization. Although the high temporal resolution and open accessibility of major spaceborne remote sensing images offer great opportunities for such purposes, their spatial resolution continues to be a limitation. Enhancing the spatial resolution of such remote sensing images using deep learning algorithms is currently a hot research topic. Multiple cases have shown that impressive results could be achieved on spaceborne optical images such as Sentinel-2 images using variants of the Generative Adversarial Networks (GAN) used in image supper-resolution, when evaluated using quantitative metrics such as signal-to-noise ratio and perceptual metrics such as structural similarity index. Practical performance of such super-resolved images in performing various purposes, compared to their comparable resolution original images, are however less researched. In this work, we investigate the relative performance of super-resolved Sentinel-2 images, in comparison to the Very High Resolution (VHR) optical multispectral images from the Copernicus Contribution Missions (CCM) and the original resolution Sentinel-2 images at 10 m spatial resolution, for detecting built-up areas. A GAN based super-resolution model is trained using ten Sentinel-2 tiles in the southern part of Norway to enhance the resolution to 2.5 m. We gathered Sentinel-2 images acquired during the summer of 2021. VHR images from the same season with spatial resolution of 2 m are obtained from the CCM. Building footprint map from the national database is acquired for the same region. A U-net type semantic segmentation model based is then trained on the three datasets separately and built-up areas are then predicted for a test region. Their results are then compared in terms of pixel accuracy, intersection over union (Jaccard Index) and the number of building footprints detected.

How to cite: Debella-Gilo, M.: Relative performance of super-resolved Sentinel-2 images for built-up area mapping using deep learning, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12801, https://doi.org/10.5194/egusphere-egu24-12801, 2024.

08:55–09:05

EGU24-17039

ECS

On-site presentation

Drones and machine learning for habitat monitoring

Charmaine Cruz, Jerome O'Connell, James R. Martin, Philip M. Perrin, Kevin McGuinness, and John Connolly

The increasing decline in the status of habitats, mainly due to anthropogenic stressors, has spurred the development and implementation of many conservation-related legislation. This legislation involves mapping, a critical task to gather important information about the habitats, including their locations, spatial extents and changes over time. Advancements in mapping technologies, such as drones and machine learning, can be integrated with conventional field surveys to improve the efficiency and effectiveness of mapping habitats. The study assessed the effectiveness of drone-acquired data and machine learning as tools for accurate and detailed mapping of highly dynamic and fine-scale mosaics of habitats in a coastal dune environment. Drone imagery and field data were acquired in a dune site in Kerry, Ireland, during the growing season in 2020: May (early), July (mid) and October (late). Topographic data representing the terrain of the site were also generated during the photogrammetric processing of the images. These datasets were then processed and analysed using the Random Forest machine learning technique to classify dune habitats at this site. The results showed that using multiple drone datasets acquired throughout the vegetation growing season achieved higher classification accuracy compared to using just a single dataset (92.37% vs. 84.09%, respectively). Also, including topographic data consistently improved the accuracy, regardless of the number of datasets. Comparing the three drone-acquired datasets, the analysis suggested that the dataset acquired in the middle period of the growing season, i.e., the flowering period, was better than those acquired in the early or late periods for dune habitat mapping.

A critical aspect of habitat conservation is tracking the location and expansion of invasive species, which is considered a major threat to and pressures on habitats. This study also explored the potential of utilising drone imagery and deep learning (DL) techniques for mapping invasive species, including those in the early stages of invasion, i.e., occurring in small patches. However, creating a robust DL model is challenging due to the requirement for large and diverse labelled training data. The study implemented a DL semantic segmentation on drone imagery and investigated the potential of applying data augmentation and pseudo-labelling to increase the amount and diversity of labelled data. Results showed that DL-based segmentation achieved high accuracy (mean Intersection-Over-Union [mIOU] score=0.832). The model trained on the augmented and pseudo-labelled data achieved an mIOU score of 0.712 on an independent dataset, while there was a decrease of 0.158 in model performance when only the original labelled data were used. This result suggests the potential of using data augmentation and pseudo-labelling techniques in creating more robust models.

Overall, the combined use of high-resolution drone data and machine learning techniques offers massive potential for repeatable and systematic approaches to fine-scale habitat characterisation. The resulting detailed maps from these approaches can provide critical information to guide and inform habitat conservation efforts. Moreover, these maps can support and contribute to the evidence-based implementation of SDG 15, which focuses on “protecting, restoring and promoting sustainable use of terrestrial ecosystems and preventing biodiversity loss”.

How to cite: Cruz, C., O'Connell, J., Martin, J. R., Perrin, P. M., McGuinness, K., and Connolly, J.: Drones and machine learning for habitat monitoring, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17039, https://doi.org/10.5194/egusphere-egu24-17039, 2024.

09:05–09:15

EGU24-5227

On-site presentation

Spatio-temporal changes in vegetation structure and its driving factors based on multi-resolution remote sensing images in arid and semi-arid regions

Jie Bai

Arid and semi-arid regions occupy 41% of the global land area and support more than 38% of the global population. With the limited precipitation and atmospheric condensation, the survival of natural vegetation here mainly depends on the shallow/deep soil water and groundwater. The natural ecological barrier formed by natural vegetation plays an important role in protecting the stability of artificial oases and maintaining ecological security in desert areas here. Based on multi-source and multi-resolution remote sensing images (MODIS, Landsat, GF-2), as well as meteorological data, Gravity Recovery and Climate Experiment (GRACE) data, and groundwater table depth (GTD) data, this study took Xinjiang, China, as a research area to accurately characterize the spatio-temporal changes of its vegetation structure (vegetation coverage, vegetation index) and quantify its driving factors. The results showed as following:

(1) At the scale of the whole of Xinjiang, the modified three-band maximum gradient difference (TGDVI) method was proposed to improve the extraction accuracy of fractional vegetation cover (FVC) in desert areas. It showed that the average FVC of Xinjiang had an increasing trend as a whole, with a growth rate of 0.19%∙a^–1 from 2003 to 2020. The influence of temperature on FVC was mainly concentrated in spring and autumn, while precipitation and groundwater storage (GWS) were the main factors in summer. In summer, GWS was the main factor affecting FVC of shrubland and cropland, and precipitation had the greatest impact on FVC of meadow grassland. With the hydrothermal conditions becoming wetter, the influence of FVC on temperature in Xinjiang gradually decreases, but that on precipitation increases. The influence of GWS on FVC increased from arid to semi-arid conditions, and then it decreased from semi-arid to humid conditions.

(2) At the scale of inland river basin, taking the Sangong River Basin (SRB) as an example, this study proposed to use deep learning methods based on GF-2 images to accurately extract the distribution of desert woody plants, and combined with Landsat 5\7\8 images to establish high-precision time series data set of woody plants vegetation index. It found that the difference between woody plant coverage (WFC) and vegetation coverage (FC) was a fine indicator for delineating desert-oasis ecotone. In this study, this range was about 1.6 km m from the oasis. Increasing GTD inhibited desert vegetation growth, while Enhanced Vegetation Index (EVI) changes in croplands were closely related to the expansion of agricultural areas rather than GTD. In the desert of lower reaches of SRB, the growth of woody plants was constrained as GTD increased.

This study indicated that groundwater was a critical influencing factor in maintaining vegetation survival in arid and semi-arid regions at both regional and watershed scales, especially for woody plants. This was of great significance to the sustainable utilization of water resources and the protection of the ecological environment here.

How to cite: Bai, J.: Spatio-temporal changes in vegetation structure and its driving factors based on multi-resolution remote sensing images in arid and semi-arid regions , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5227, https://doi.org/10.5194/egusphere-egu24-5227, 2024.

09:15–09:25

EGU24-8657

ECS

Highlight

On-site presentation

An innovative SAM-ViT based tool for the automatic detection of litter items on sandy beaches

Angelo Sozio, Angela Rizzo, Vincenzo Mariano Scarrica, Pietro Patrizio Ciro Aucelli, Giorgio Anfuso, Giovanni Barracane, Luca Antonio Dimuccio, Rui Ferreira, Marco La Salandra, Antonino Staiano, Maria Pia Tarantino, and Giovanni Scicchitano

Machine learning (ML) techniques in the field of Computer Vision turned out to be well-performing tools for the automatic detection of beach litter (BL) items on high resolution UAVs images. This study was carried out in the frame of the RiPARTI project (funded by Apulia Region) proposes an innovative approach based on the combination of the aero-photogrammetric surveys with a newly proposed ML tool.

A series of experiments were conducted with a Mask-Regional Convolutional Neural Network-based (Mask-RCNN) algorithm using an image dataset acquired UAVs fights performed on different coastal sites, in Italy, Portugal, and Spain. Preliminary detection experiments were conducted using three BL items categories, “Bottles”, “Worked Wood” and “Nets”. Subsequently, a comparison with algorithms available in QGIS software confirmed the great potential of Computer Vision techniques. Indeed, in previous studies (Sozio et al., 2023), the performance of the Mask-RCNN based algorithm resulted higher than performances of algorithms available in QGIS software, but still not enough to obtain a definitive ML tool for BL automatic detection.

The novel ML tool here proposed exploits the powerful dataset of Segment Anything (SAM) (Kirillov et al., 2023) developed by Meta AI, as segmentation algorithm and Visual Transformer (ViT) for the classification task. A first experiment was conducted with a dataset derived from UAVs images acquired in five different sites, i.e., Capitolo and Torre Guaceto beach (Italy), Leirosa beach (Portugal), Valdelagrana (Spain), and Cala del Cefalo beach (Italy). Aero-photogrammetric surveys were carried out at different flight heights for each site so, the final images resolution ranges from 0.3 cm/pixel to 0.7 cm/pixel. Moreover, the different color of the sand (background) represents a parameter which could affect the performance of segmentation process. Orthomosaics in .tiff format were split in 1000-pixel square tile and segmented by SAM. It executed a panoptic segmentation that produced 450 masks, both concave and convex, corresponding to objects identified on images. These masks were catalogued according to 11 labels (Bottles, Nets, Polystyrene, Worked wood, Vials, Buckets, Building waste, Ethernit, Sand, Vegetation, and Water), accounting for both the most common litter categories and natural assets. Subsequently, masks so gathered were used to train ViT, the classification algorithm and to perform the test phase, which was carried out on 450 masks, with a ratio of training, validation and test split of 7/10, 1/10 and 2/10, respectively. A preliminary experiment produced output images classified by ViT with an accuracy of 0.93 and an f1 score equal to 0.6. Data considered for this last experiment are more complex for number of classes and amount of data, so performance are better in projection” also considering the different images resolution and the background texture. Finally, identified items are georeferenced with a projected reference system. The method outstanding a very reliable performance for the BL detection task and could represent a useful and definitive approach for the assessment of the BL distribution and as well as for the identification of the main accumulation zones so as to make possible the development of tailored coastal management actions.

How to cite: Sozio, A., Rizzo, A., Scarrica, V. M., Aucelli, P. P. C., Anfuso, G., Barracane, G., Dimuccio, L. A., Ferreira, R., La Salandra, M., Staiano, A., Tarantino, M. P., and Scicchitano, G.: An innovative SAM-ViT based tool for the automatic detection of litter items on sandy beaches, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8657, https://doi.org/10.5194/egusphere-egu24-8657, 2024.

09:25–09:35

EGU24-21514

Highlight

On-site presentation

Monitoring land surface temperature from space -constellr HiVE - new perspectives for environmental monitoring

Daniel Spengler, Elsy Ibrahim, Nicolas Chamberland, Ariadna Pregel Hoderlein, Jonas Berhin, Tianran Zhang, and Matthieu Taymans

The importance of thermal remote sensing satellite data has become increasingly recognised for environmental monitoring in the recent years. Specifically, in providing valuable information on the earth’s land surfaces and plant canopies, which can be used to diagnose water stress and drought conditions. Thermal data, with its unique capacity to capture temperature variations, brings an unexplored added value for understanding processes. Thermal imaging also provides much-needed information on soil moisture status in regions where data are scarce. The use of thermal remote sensing data has been widely recognized as a contributor to large-scale environmental monitoring. It has enabled getting more insights from a variety of resource management domains, including agriculture, forestry, geological resources, water resources, cryosphere, atmosphere, and analytics for climate change.

Unfortunately, current thermal satellite data are either only available in very high to high temporal resolution or in low spatial resolution (>1,000m) respectively from geostationary or sun-synchronous satellites, which have a low temporal resolution and moderate spatial resolution of 30-100m. Such data are therefore only of limited suitability when regular monitoring of small-scale environmental factors is required.

constellr develops a constellation of new state-of-the-art satellites covering the visible, near infrared, and thermal infrared parts of the spectrum at high-resolution, planned for launch by the end of 2024. The HiVE (High-precision Versatile Ecosphere monitoring mission) constellation comprises micro-satellites in the 100 kg class, orbits in a sun-synchronous plane at an altitude of 550 kilometers. With a remarkable 1-day global temporal resolution (5 sat from 2026), 30 meters spatial resolution, and up to 1.5 K absolute temperature accuracy, HiVE is uniquely equipped with a cryo-cooled thermal detector to provide accurate and timely data for environmental monitoring. Leveraging its proprietary data, imagery from public missions, and strong remote-sensing expertise for data fusion, harmonization, and analytics, constellr offers timely and highly scalable solutions.

To ensure the interoperability of HiVE data with existing and upcoming thermal satellite missions, constellr has developed a proprietary LST retrieval algorithm. Based on Top of Atmosphere (TOA) radiance and atmospheric condition, emissivity and temperature estimations are performed. This enables capturing the vegetation dynamic related to specific vegetation types and growth stages. To further reinforce its monitoring capacity constellr leverages existing thermal infrared sensors and performs a spatial and spectral harmonization of the data to provide a consolidated LST product at 30m resolution (LST30).

For the HiVE satellites we will present the mission concept, status, and planned validation activities. Furthermore, the added value of high resolution LST data at 30m will be presented with diverse use cases.

How to cite: Spengler, D., Ibrahim, E., Chamberland, N., Pregel Hoderlein, A., Berhin, J., Zhang, T., and Taymans, M.: Monitoring land surface temperature from space -constellr HiVE - new perspectives for environmental monitoring, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-21514, https://doi.org/10.5194/egusphere-egu24-21514, 2024.

09:35–09:45

EGU24-14048

ECS

On-site presentation

Integrating Deep Learning and the Google Earth Engine Could Computing Platforms for Fast and Accurate Cloud Masking

Chengkang Zhang, Yueming Wang, Wen Nie, Yunliang Qi, Lei Zhang, and Lei Ren

In dealing with optical satellite images, accurate and efficient cloud positioning and masking are often the prerequisites for the subsequent tasks. However, the current cloud masking algorithms have difficulty in achieving these two goals at the same time. On the one hand, though many physically based models such as the Fmask algorithms have been proposed and widely applied, the performance of these methods still has some limitations including manually adjusted parameters in selecting the proper index and poor performance in thin cloud detection. This situation is much worse when applying the built-in Fmask in cloud-computing platforms such as the Google Earth Engine (GEE). On the other hand, an increasing number of sophisticated algorithms based on Deep Learning such as Convolutional Neural Networks (CNN) and Transformers have been proposed, they are, in most cases, deployed in local environments and require huge amounts of computational capacity to accomplish the tasks, which is inefficient and cannot be quickly utilized in large-scale and long-term studies, especially shows limitations with the trial of transferring the model into the GEE platforms. To solve the aforementioned dilemmas, the present study proposes a novel method for cloud masking by integrating deep learning and cloud-computing GEE. First, we construct a cloud dataset that is composed of globally selected cloud-contaminated pixels with the Fmask algorithm. Then, we use this cloud dataset to train a lightweight and flexible deep learning model based on LeNet. Last to screen out cloud pixels. Last, the developed model is transferred into the cloud-computing GEE platform and used to conduct cloud masking for each optical satellite image. The results show that in comparison to the conventional Fmask algorithm, the performance of the proposed exists superior in both detecting thick and thin clouds. More importantly, cloud masking can be achieved without bureaucratic procedures such as first downloading images and uploading the cloud masking, which is often required by locally developed deep learning models. By utilizing the proposed method, accurate and fast cloud detection can be achieved on the GEE platform that can be used in subsequent tasks like image compositing. For example, the generated monthly mean composting images show much better performance in visualizing ground objects when compared to those images based on the Fmask method, as the remaining cloud pixels that cannot be detected by the built-in Fmask algorithms can be more accurately examined. Through the usage of the proposed cloud mask methods, the merits of the powerful strength in data fitting and model optimization of deep learning algorithms and high efficiency in dealing with big data of GEE are naturally integrated, which is expected to shed light on cloud masking and other remote sensing modeling tasks.

How to cite: Zhang, C., Wang, Y., Nie, W., Qi, Y., Zhang, L., and Ren, L.: Integrating Deep Learning and the Google Earth Engine Could Computing Platforms for Fast and Accurate Cloud Masking, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14048, https://doi.org/10.5194/egusphere-egu24-14048, 2024.

09:45–09:55

EGU24-15432

ECS

On-site presentation

Ordinary and Bundle-Unmixing Approaches Under the Influence of the BRDF Effect

Fadi Kizel

Remotely sensed spectral data play a vital role in interpreting and understanding land cover properties. However, due to its typically low spatial resolution, the ability to extract information from the data using traditional applications is limited to the pixel size, which, in many cases, is bigger than the phenomena of interest. On the other hand, spectral unmixing allows for extracting information from such data at the subpixel level by estimating the fractional abundance of the different land cover types within the pixel, so-called endmembers (EMs). Many approaches have been developed for this purpose. Nonetheless, the ordinary methods use only a single spectral signature for each EM, disregarding the highly probable spectral variability within each EM spectra. Therefore, bundle-unmixing methods were developed to overcome this limitation by using a set of spectra for each EM.Previous research results show the advantage of bundle-unmixing methods in enhancing the fraction estimation for various cases with an EM variability due to multiple effects. Still, despite the encouraging results, only very few works considered the spectral variability caused by the impact of the Bidirectional Reflectance Distribution function (BRDF). Thus, this work focused on studying the performance of ordinary and bundle-unmixing methods under the influence of the BRDF effect. We comparatively examined five methods, each with particular characteristics: two bundle methods and three ordinary ones; among them, one method relies on the Spectral Angle Mapper (SAM) as an objective function. We used three data sets for experimenting: 1) a laboratory set involving three land covers measured from various viewing zenith angles, 2) a synthetic data set created by simulating spectral data influenced by the BRDF effect using semi-empirical models, and 3) an areal data set including hyperspectral images over the Icelandic volcanic area. We examined the methods' performance under various signal-to-noise (SNR) ratio levels.The results clearly show the superiority of the bundle methods in reducing the effect of the BRDF on the estimated fractions. Besides, the most exciting outcome shows that despite relying on only one spectral signature per EM, the SAM-based method outperforms the other ordinary methods and provides accurate results comparable to the bundle ones. Our study aimed to understand better the complex relationship between BRDF, spectral variability, and unmixing accuracy. This investigation may enhance remote sensing data analysis and refine the unmixing approach in the presence of BRDF-induced spectral variability to improve land cover mapping and environmental monitoring using spectral data.

How to cite: Kizel, F.: Ordinary and Bundle-Unmixing Approaches Under the Influence of the BRDF Effect, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15432, https://doi.org/10.5194/egusphere-egu24-15432, 2024.

09:55–10:05

EGU24-9836

On-site presentation

Research and Development of Satellite-borne Scanning Array for Hyper-multispectral Radiowave Imaging (SAMRAI)

Takashi Maeda, Yuta Kobayashi, Nguyen Tat Trung, Tsutomu Yano, and Naoya Tomii

Scanning Array for hyper-Multispectral RAdiowave Imaging (SAMRAI) is a passive interferometric radiometer. In this respect, it is similar to MIRAS on board the SMOS satellite launched by ESA, but it realizes ultra-wideband (1-41 GHz) and high-frequency-resolution (27 MHz) microwave spectrum measurement. We believe that SAMRAI is the world's first microwave hyperspectral radiometer.

JAXA will continue to operate the satellite-borne microwave radiometer AMSR series for more than 30 years, including AMSR3 currently under development. The design has remained largely unchanged for 30 years, and various issues are becoming apparent. In particular, the radio frequency interference (RFI) contaminating the natural-origin signals is a serious problem, and we believe that microwave hyperspectral measurement is essential for identifying and isolating RFI signals. This was a big motivation for developing SAMRAI. However, microwave hyperspectral measurement must have new possibilities, such as making it possible to measure the frequency characteristics of the emissivity of the Earth surface.

SAMRAI's key technologies are:
1) ultra-highspeed A/D conversion equivalent to 80 GSPS (SPS : samples per second)
2) ultra-wideband antenna (1 - 41 GHz) as an element of a phased array antenna (PAA) system
3) on-board data processing for an interferometer using FPGA including mitigation of grating lobes

SAMRAI was first developed to be mounted on a helicopter (aircraft-borne SAMRAI), and its performance is currently being confirmed. Based on performance confirmation of the aircraft-borne SAMRAI, the final goal is to develop the satellite-borne SAMRAI and launch it in 2027.

Here, we presents the detail of the technical points of SAMRAI, the performance confirmation results of the observation experiment by the aircraft-borne SAMRAI and investigation of a new algorithm for geophysical value retrieval based on microwave spectrum measurement.

How to cite: Maeda, T., Kobayashi, Y., Trung, N. T., Yano, T., and Tomii, N.: Research and Development of Satellite-borne Scanning Array for Hyper-multispectral Radiowave Imaging (SAMRAI), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9836, https://doi.org/10.5194/egusphere-egu24-9836, 2024.

10:05–10:15

EGU24-15649

Highlight

On-site presentation

TRISHNA : towards innovative infra-red data for environmental monitoring

Jean-Louis Raynaud, Philippe Gamet, Corinne Salcedo, Loïc Lyard, Carole Amiot, Cécile Picard, Jean-Louis Roujean, and Bimal Bhattacharya

Planned to be launched in 2026, the TRISHNA mission (Thermal infraRed Imaging Satellite for High-resolution Natural resource Assessment) aims at providing visible and infra-red measurements with a worldwide coverage, for continents and coastal ocean. During its 5-year mission lifetime, it will acquire signal of the surface-atmosphere system with a 60-meter resolution. With such characteristics, TRISHNA mission can be considered as a precursor for future missions, such as SBG (NASA) or LSTM (ESA/COPERNICUS), with which a synergy has been established to allow a long term data inter-comparison.

Resulting from a cooperation between CNES and ISRO, TRISHNA mission will take advantage of two mission centers, in France and in India, to provide users with Level 1C, 2 and 3 products. Based on identical algorithms specifications and products definitions, these ground segments will process and deliver data with Near Real Time demonstration objectives. Indeed, to allow quick decisions, some Level 2 products (for instance water stress indexes) shall be made available to users in less than 12 hours after acquisition.

This presentation will give an update of the status of TRISHNA mission, focusing on the details of products which will be provided and their potential application in environmental remote sensing. The French ground segment will also be described to enhance how TRISHNA data will be processed and delivered.

How to cite: Raynaud, J.-L., Gamet, P., Salcedo, C., Lyard, L., Amiot, C., Picard, C., Roujean, J.-L., and Bhattacharya, B.: TRISHNA : towards innovative infra-red data for environmental monitoring, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15649, https://doi.org/10.5194/egusphere-egu24-15649, 2024.

Posters on site: Wed, 17 Apr, 16:15–18:00 | Hall X4

Display time: Wed, 17 Apr 14:00–Wed, 17 Apr 18:00

Chairpersons: Gaetana Ganci, Gabor Kereszturi

X4.165

EGU24-78

ECS

TFe and SiO2 Spatial Mapping Enhancement in Iron Tailings: Efficiency of a Calibration Transfer Model

Nisha Bao, Haimei Lei, Yue Cao, Asa Gholizadeh, Mohammadmehdi Saberioon, and Yi Peng

Iron tailings are crystalline powders predominantly composed of iron (Fe) and silicon dioxide (SiO₂). Spatially characterizing of the physical and chemical properties of iron tailings is greatly important for optimal utilization and proper disposal of tailings. Visible–near infrared–shortwave infrared (VIS–NIR–SWIR; 350–2500 nm) spectroscopy offers a rapid, non-destructive, and cost-effective method for quantitatively analyzing tailings properties. The main objective of this study was to map the spatial distribution of total Fe (TFe) and SiO₂ content in a tailings dam through the use of laboratory spectra and GF-5 hyperspectral imagery based a calibration transfer model approach. A total of 77 samples were collected from the surface of targeting field and scanned by a laboratory VIS–NIR–SWIR reflectance spectrometer. The competitive adaptive re-weighted sampling (CARS) algorithm was applied to select important spectral features. Subsequently, different spectral indices were calculated to enhance the prediction performance of the calibration models. Rulefit and random forest (RF) algorithms were used to calibrate spectral information with associated tailing properties. The results showed that the Rulefit algorithm with selected feature bands and calculated spectral indices yielded the highest estimation accuracy for TFe (R² = 0.86, RMSE = 1.30%, LCCC = 0.87 and bias = -0.45) and SiO₂ (R² = 0.74, RMSE = 2.00%, LCCC = 0.84 and bias = 0.38). The direct standardization (DS) algorithm was applied to correct GF-5 hyperspectral images and enhance the efficiency of calibration model transfer process. Finally, the Rulefit models were transferred to corrected GF-5 hyperspectral images for mapping the spatial distribution of TFe and SiO₂ contents. Our results demonstrated the possibility of successful transfer of laboratory spectral-based model to the GF-5 hyperspectral imagery for mapping spatial distribution of tailing compositions. This finding can be applied for efficiently recovering valuable metals and minimizing environmental risks.

How to cite: Bao, N., Lei, H., Cao, Y., Gholizadeh, A., Saberioon, M., and Peng, Y.: TFe and SiO2 Spatial Mapping Enhancement in Iron Tailings: Efficiency of a Calibration Transfer Model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-78, https://doi.org/10.5194/egusphere-egu24-78, 2024.

X4.166

EGU24-151

Satellite Images Analysis for Sargassum Recognition

(withdrawn)

Rodrigo Montufar-Chaveznava, Jesus Daniel Vilchis-Ibarra, and Ivette Caldelas-Sanchez

X4.167

EGU24-310

Estimating soil surface moisture in areas with high fire incidence in Serra da Canastra National Park, Brazil, using OLI and TIRS landsat-8 sensor data

Marcos César Ferreira and Danilo Carneiro Valente

The Serra da Canastra National Park (SCNP) is an important conservation unit of the Cerrado biome (Brazilian savanna) located in a mountainous area in southeastern Brazil, where the headwaters of the São Francisco River, one of the longest rivers in the country, are located. In this area, large forest fires have occurred annually, destroying native vegetation and fauna. Due to the large expanse of the Serra da Canastra National Park, firefighting in this location is a very difficult and time-consuming task. This article presents a methodological procedure for mapping soil surface moisture in areas that have shown high fire frequencies over a timespan longer than 20 years in the national park using images from the OLI (Operational Land Imager) and TIRS (Thermal Infrared Sensor) instruments onboard the Landsat-8 satellite. A soil surface moisture map was generated based on the drying of the vegetation using the temperature index calculated from a scatterplot of surface temperatures and vegetation index values in each pixel image. The vegetation index values were calculated from the normalized difference vegetation index (NDVI) algorithm using the red and near-infrared spectral bands of the OLI. The soil surface temperature was estimated using data from the TIRS sensor. We observed that areas with minimal moisture occurred on convex slopes, while areas with maximal moisture occurred on concave slopes. In addition, we found that areas with minimal moisture were more frequently located on north- and northeast-facing slopes, while areas with maximal moisture were more frequently found on south- and southwest-facing slopes. The accuracy of the soil surface moisture map was evaluated based on relative volumetric soil moisture data collected in the field using an instantaneous electronic soil moisture meter. The results of this research may contribute to the monitoring and forecasting of other areas highly susceptible to the occurrence of fire and to the planning of firefighting actions in the Serra da Canastra National Park.

How to cite: César Ferreira, M. and Carneiro Valente, D.: Estimating soil surface moisture in areas with high fire incidence in Serra da Canastra National Park, Brazil, using OLI and TIRS landsat-8 sensor data, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-310, https://doi.org/10.5194/egusphere-egu24-310, 2024.

X4.168

EGU24-319

ECS

OSPP

A remote sensing index for assessing long-term ecological impact in arid mined land

Dantong Meng, Nisha Bao, Tianhong Yang, and Qiyue Li

Satellite remote sensing technology, with its ability to record spatial and temporal land surface conditions, has been extensively and effectively utilized in evaluating mining environments. Western China, characterized by arid ecosystems, is a significant mineral resource area, boasting at least ten super-large mineral bases, including coal, non-ferrous metal ores, and metal mines. Surface mining activities, marked by large spatial and temporal scales, can exacerbate the fragility and changes to the ecological environment in vulnerable areas. Consequently, it is crucial to assess and comprehend the spatial and temporal impacts of mining on arid ecological systems for green mine construction and mine reclamation. This study focuses on three typical open-pit mines in the arid regions of Xinjiang, China (site Ⅰ: Jinbao Iron open-pit mine, site Ⅱ: Heishan coal open-pit mine, site Ⅲ: Wulagen Lead Zinc open-pit mine). The primary objective was to develop a remote sensing index (Mined Land Ecological Status Index, MLESI) that considers biological factors such as dryness, bare soil flatness, land surface temperature, and slope to assess the ecological status in arid mining areas. Subsequently, Principal Component Analysis was employed to couple these four factors to construct MLESI. The efficacy of MLESI was compared with the Remote Sensing Ecological Index (RSEI) and the Land Surface Ecological Status Composition Index (LSESCI) in different landforms in arid mining areas characterized by bare soil and rock. The spatial and temporal changes in mining effects from 2005 to 2020 were analyzed using the Sen+Mann-Kendall method based on Landsat time series images. The results indicated that the average Pearson correlation coefficient (r) between MLESI and each factor exceeded 0.65. The heat factor had the highest correlation coefficient of 0.8 with MLESI for mine site Ⅰ, while the dryness factor had the highest correlation of 0.82 with MLESI for mine site Ⅱ. The slope factor had the highest correlation coefficient of 0.82 for mine site Ⅲ. For mine sites Ⅰ and Ⅱ, the LSESCI overestimated the areas of poor ecological status, identifying most of the natural land as poor. RSEI was unable to reveal the changes in ecological status correlating with landform variety. In general, MLESI was not only highly effective in characterizing the mining area from natural bare soil and rock lands but also in indicating ecological changes along the mining direction with landform changes. The ecological status in mine site I deteriorated by 54.84% since 2015 due to extensive surface mining activities. For mine site II, the ecological status gradually declined from an MLESI value of 0.68 in 2005 to 0.38 in 2020, with a total of 2.36 km2 area experiencing significant changes to poor ecological status over 15 years. For mine site Ⅲ, the ecological status improved due to land reclamation, reaching the highest MLESI value of 0.77 in 2017. An area of 0.95 km2 experienced significant changes to good ecological status from 2008 to 2020. Therefore, the proposed MLESI outperformed RSEI and LSEISCI in monitoring different mine ecological statuses in a typical arid ecosystem.

How to cite: Meng, D., Bao, N., Yang, T., and Li, Q.: A remote sensing index for assessing long-term ecological impact in arid mined land, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-319, https://doi.org/10.5194/egusphere-egu24-319, 2024.

OSPP judging

X4.169

EGU24-484

ECS

OSPP

Spatiotemporal fusion of multi-source Chinese Gaofen remote sensing data for mapping costal wetland using stacking ensemble classification method

Yitong Liu, Nisha Bao, Huiya Qian, and Dantong Meng

The precise mapping of coastal wetlands holds great significance in the context of monitoring carbon sequestration and storage within coastal ecosystems, particularly in light of climate change and human-induced activities. Time series and multi-source remote sensing data offers distinct advantage in the spatial and temporal mapping of land use, particularly in wetland systems, encompassing various types of vegetation. The primary aim of this study was to delineate the spatial distribution of land use within the Liao River Delta (LRD) wetland. This was achieved by employing a stacking ensemble model that integrated time series GaoFen-1 (GF-1) optical imagery, GaoFen-3 (GF-3) synthetic aperture radar (SAR) imagery, and Moderate Resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) imagery. The first step involved the application of the enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM) to fuse the GF-1 NDVI and MODIS NDVI datasets, resulting in the production of time series NDVI data. Subsequently, the parameters pertaining to vegetation phenology were obtained by employing the threshold method on time series NDVI data. We compiled feature datasets that encompassed GF-1 spectral bands, spectral indices, phenological parameters, and GF-3 SAR features. In order to mitigate data redundancy, the Recursive Feature Elimination and Cross-Validation (RFECV) model was employed to identify and select significant features. Finally, the stacking ensemble model was constructed by combining five base models [K-Nearest Neighbors (KNN), Random Forest (RF), Adaptive Boosting (AdaBoost), Extreme Gradient Boosting (XGBoost), and Light Gradient Boosting Machine (LightGBM)] to perform wetland classification. The findings showed the following: (1) ESTARFM was able to successfully fuse GF-1 NDVI and MODIS NDVI data, resulting in a spatiotemporal fusion with a coefficient of determination (R²) of 0.85 and a root mean square error (RMSE) of 0.07. (2) In the process of wetland classification, the RFECV algorithm was employed to select relevant features. Specifically, 75 spectral band features, 89 spectral index features, 13 SAR features, and 7 phenological parameters were identified as significant for this task. (3) A stacking ensemble model was constructed using the aforementioned multi-source features. This model exhibited a robust and consistent performance in wetland classification, achieving the highest overall accuracy of 94.33%. Notably, this accuracy improvement ranged from approximately 0.09% to 10.02% when compared to the individual base models. Thus, the present study has the potential to be utilized in the context of fine-scale wetland monitoring, thereby offering valuable assistance to the field of wetland environmental research.

How to cite: Liu, Y., Bao, N., Qian, H., and Meng, D.: Spatiotemporal fusion of multi-source Chinese Gaofen remote sensing data for mapping costal wetland using stacking ensemble classification method, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-484, https://doi.org/10.5194/egusphere-egu24-484, 2024.

OSPP judging

X4.170

EGU24-650

ECS

Quantitative the urban vegetation carbon storage by constructing seamless and dense time series remote sensing data

(withdrawn)

Ya Zhang, Zhenfeng Shao, Xiaodi Xu, and Qingwei Zhuang

X4.171

EGU24-4282

ECS

Atmospheric correction method and application of satellite hyperspectral data: A case study in mineral resource monitoring

Qianqian Jiang, Tao He, and Wenchang Li

Hyperspectral sensors have become indispensable tools in remote sensing applications, playing a pivotal role in precision agriculture, mineral alteration mapping, land cover classification, and autonomous satellite wildfire detection. Their high spectral resolution and comprehensive spectral information about ground objects underscore their significance. However, despite the wealth of hyperspectral data sources, challenges in data processing have impeded their quantitative application on the Earth's surface. Atmospheric correction, a critical procedure for deriving surface reflectance that accurately captures surface properties, is essential to overcoming these challenges.

In this context, we present an adaptive and automated atmospheric correction scheme specifically designed for hyperspectral data. This approach involves the inversion of atmospheric parameters, including aerosol optical depth (AOD) and precipitable water vapor content (PWV), utilizing the inherent hyperspectral information in the image. Importantly, the atmospheric correction is performed without the need for simultaneous atmospheric and surface observations. AOD retrievals exhibit excellent consistency with ground measurements (RMSE < 0.05, R² > 0.78), and PWV retrievals are also accurate, with an RMSE less than 0.17 g/cm² and an R² greater than 0.94. The acquired surface reflectance demonstrates a remarkable resemblance in terms of shape (spectral angle < 2.2°) and magnitude (RMSE < 0.02) compared to in-situ measurements.

The ZY-1 02D Satellite (ZY02D), led by the Ministry of Natural Resources of China, is purposefully designed for monitoring natural resources. The Advanced HyperSpectral Imager (AHSI) aboard ZY02D covers a wide area of 60 km * 60 km with a medium-to-high resolution of 30 m and a spectral range spanning from 400 to 2500 nm, featuring 166 channels. To monitor mineral resources in Guangdong Province, China, ZY02D AHSI images are acquired. After atmospheric correction using our method, the resulting surface reflectance spectral curve is utilized to identify mineral areas. The similarity between satellite-based and in-situ measured surface reflectance is assessed by spectral angle and Euclidean distance to identify potential mineral areas. The absorption characteristics of minerals are extracted from the satellite-based surface reflectance to enhance the results. Through comparison with mining rights maps, more than 80% of mining area maps are successfully identified. Moreover, it facilitates the monitoring of unauthorized mining activities, thereby improving the enforcement efficiency of government agencies.

How to cite: Jiang, Q., He, T., and Li, W.: Atmospheric correction method and application of satellite hyperspectral data: A case study in mineral resource monitoring, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4282, https://doi.org/10.5194/egusphere-egu24-4282, 2024.

X4.172

EGU24-4513

ECS

Measurement of Atmospheric CH4 by 1.6 μm High Resolution Non-Modulated Laser Heterodyne Radiometer (NM-LHR)

Jun Huang, Wenyue Zhu, Yinbo Huang, Zhensong Cao, Xingji Lu, and Yuan Meng

Laser heterodyne spectroscopy detection technology boasts exceptional advantages such as high spectral resolution and high signal-to-noise ratio (SNR). It excels at capturing spectral line broadening information of upper atmospheric molecules. This presents substantial research value in the realms of greenhouse gas profile measurement and the assessment of laser propagation effects in the atmosphere. This paper delves into the investigation of the processing method for heterodyne signals, adopting a non-modulated signal processing approach to construct a near-infrared non-modulated fiber laser heterodyne radiometer. This innovative design significantly enhances the device's response speed and SNR. The radiometer achieves a spectral resolution of 0.006 cm^-1 and an SNR of 300. This facilitates the acquisition of vertical profile distribution and column concentration of CH₄ by measuring the absorption line of atmospheric CH₄. Comparative tests reveal compelling advantages of the non-modulated device, with the modulated device collecting data 6 times in 6 minutes, yielding an SNR of 58. In contrast, the non-modulated device demonstrates superior efficiency by collecting data 6000 times in 2 minutes, resulting in a remarkable SNR of 103. The inversion results of CH₄ column concentration from the laser heterodyne radiometer were compared with those from the Fourier transform spectrometer (EM27/SUN), with average concentrations of 1.88×10^-6 and 1.93×10^-6, exhibiting an overall deviation of approximately 2.6%. The non-modulated laser heterodyne radiometer provides a new reference for the rapid, accurate and high spectral resolution measurement of greenhouse gas concentration.

How to cite: Huang, J., Zhu, W., Huang, Y., Cao, Z., Lu, X., and Meng, Y.: Measurement of Atmospheric CH4 by 1.6 μm High Resolution Non-Modulated Laser Heterodyne Radiometer (NM-LHR) , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4513, https://doi.org/10.5194/egusphere-egu24-4513, 2024.

X4.173

EGU24-5257

A Framework for Monitoring Above Ground Biomass of Hyper-Arid Rangelands in the Middle East

Kasper Johansen, Jorge Rodriguez Galvis, Hua Cheng, Samer Almashharawi, and Matthew McCabe

In line with Saudi Vision 2030 and the ambitions of the Saudi Green Initiative to plant 10 billion trees and protect 30% of Saudi Arabia’s land and sea areas, large parts of Saudi Arabia are being transitioned into nature reserves and undergoing regreening activities, while also fencing off large areas for vegetation restoration and protection. To effectively manage the regreening and restoration initiatives, it is imperative to frequently monitor biomass changes over time and quantify the accumulation of biomass to determine if the restoration and tree-planting initiatives have the intended outcomes. However, landscape responses to regreening and vegetation restoration are poorly understood in hyper-arid rangelands. Environmental processes within different habitat zones might differ, which further complicates biomass monitoring. Here, we present a framework that is currently being applied across multiple hyper-arid rangelands in Saudi Arabia to estimate biomass within newly established nature reserves. The initial work focused on developing a scaling approach between ground, unmanned aerial vehicle (UAV), and satellite image data, including PlanetScope and Sentinel-2 imagery. Field sites were identified using Google Earth imagery and a number of criteria, including a 5-year Sentinel-2 NDVI time-series to identify greening events, terrain characteristics based on DEM data, and differences in vegetation functional types (annual and perennial grass/herbs/forbs, shrubs and trees), soil types, and habitats. Field-based measurements at selected sites focused on determining biomass of different vegetation functional types, using a double-sampling approach, including a limited number of destructive samples, and a large number of biomass estimates based on the structural and dimensional characteristics of the destructive samples. UAV-based light detection and ranging (LiDAR) and multispectral data were obtained for each site to estimate biomass from the field-based samples using different machine and deep learning approaches (random forest, support vector machines, vision transformer, UNet with attention encoder). It was found that LiDAR data provided useful information on vegetation height and volume, which improved the biomass estimates, whereas the multispectral data enabled discrimination between photosynthetic and non-photosynthetic vegetation components. Based on the UAV-derived estimates of biomass, a scaling approach was applied to estimate biomass from both PlanetScope and Sentinel-2 data, with results demonstrating that the higher spatial resolution of the PlanetScope data improved the accuracy due to the very sparse vegetation in most parts of the hyper-arid rangelands of this study. While hyper-arid rangelands are generally underrepresented in research studies worldwide, this research provide a viable framework for assessing vegetation dynamics of biomass both seasonally and annually.

How to cite: Johansen, K., Galvis, J. R., Cheng, H., Almashharawi, S., and McCabe, M.: A Framework for Monitoring Above Ground Biomass of Hyper-Arid Rangelands in the Middle East, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5257, https://doi.org/10.5194/egusphere-egu24-5257, 2024.

X4.174

EGU24-7524

Introduction to Earth Observation Images and Preprocessing Operational Issues of KARI's Geostationary Satellites

Eun-Bin Park and Sang-Cherl Lee

The geostationary satellites operated by the Korea Aerospace Research Institute (KARI), known as the Cheonlian satellite series, include three satellites dedicated to Earth observation missions. Cheonlian 1 (Communication Ocean and Meteorological Satellite; COMS), with a decade-long operation, utilized two onboard instruments (MI, GOCI) for Earth observation missions, accumulating approximately 600,000 images over 10 years. The Earth observation images from Cheonlian 1 can be obtained free of charge through Korean data utilization organizations (National Meteorological Satellite Center, Korea Ocean Satellite Center). The KARI ground station holds complete data related to satellite control and operation, including the Earth observation images from Cheonlian 1. After completing its Earth observation mission, Cheonlian 1 is currently operating only its communication payload. The plan is to continue operations until the launch of Cheonlian 3 in 2027, which will take over communication payload duties. Cheonlian 2A (Geo-KOMPSAT-2A; GK-2A), launched in December 2018, succeeded Cheonlian 1 in meteorological payload Earth observation tasks. Equipped with the Advanced Meteorological Imager (AMI), developed by HARRIS, Cheonlian 2A observes the globe with 16 optical bands, focusing on the Korean Peninsula. Real-time distribution of global observation images in UHRIT, HRIT, and LRIT formats is available. Cheonlian 2B (Geo-KOMPSAT-2B; GK-2B), operated since February 2020, simultaneously operates GOCI-Ⅱ and GEMS instruments for Earth observation missions. These two instruments observe only during daytime unlike meteorological payload. GOCI-Ⅱ performs Earth observation missions with a more diverse set of optical bands compared to GOCI on Cheonlian 1. Utilizing hyperspectral sensors, GEMS observes the Earth in five different areas, enabling more accurate observation of atmospheric characteristics such as aerosols. This paper provides a detailed overview of the observation areas, data types, and validation materials of Earth observation images from KARI's geostationary satellites. Additionally, it will be introduced operational issues related to Earth observation image preprocessing system of currently operational Cheonlian 2A/2B satellites. Operational issues of image preprocessing systems include various events such as failure of image reception due to environmental factors such as Sun and radio interference. Image processing failures, such as software errors, are also part of these problems. We also discuss payload anomalies that occurred during the post-launch stabilization phase. KARI is committed to continuous research and development for the stable operation of geostationary satellites, thereby ensuring the best performance for Earth observation missions.

How to cite: Park, E.-B. and Lee, S.-C.: Introduction to Earth Observation Images and Preprocessing Operational Issues of KARI's Geostationary Satellites, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7524, https://doi.org/10.5194/egusphere-egu24-7524, 2024.

X4.175

EGU24-11264

ECS

OSPP

Observations of Changes in Vineyards Using Long-Term MODIS NDVI and LAI around Geothermal Power Plants

Merve Aydin, Michael Leuchner, and Burcak Kaynak

Turkey is ranked 4^th worldwide in terms of installed geothermal power plant (GPP) capacity (IRENA, 2023). On the other hand, there are some discussions regarding their impacts on valuable economic agricultural products like figs, olives, and grapes in intense GPP regions. Satellite-based NDVI and LAI data were used in several studies for agricultural vegetation monitoring, management and yield prediction in addition to forest vegetation. FAO (2023) utilizes monthly NDVI anomaly maps to investigate density and health of agricultural vegetation, especially for March-October covering vegetation period and phenological stages of viticulture in the Northern Hemisphere.

MODIS instruments on Terra and Aqua continuously collect NDVI and LAI data every 1-2 days in 36 spectral channels with global coverage. However, they have different spatial and temporal resolutions. Both have been used in several studies regarding vineyards. This study aims to observe changes in satellite-based NDVI and LAI for vineyards around GPPs in Turkey using several statistical analysis methods. Within this scope, the vineyard NDVI and LAI time series were obtained from 250m spatial, 16 days temporal resolutions NDVI and 500m spatial, 4 days temporal resolutions LAI between 2002-2023. All vineyard areas in the study region (AA) were defined using CORINE Land Cover (CLC) data. Three possible impacted (IA1-IA3) and two non-impacted vineyard areas (NIA1-NIA2) were selected considering prevailing wind directions and wind speeds, the capacity of GPPs, and the possible impact distances of emissions from the plants: one of which is the all-impacted area (AIA) consisting of IA1-IA3 within close proximity to GPPs and two of which are NIA1-NIA2. The correlations of NDVI and LAI were analyzed by the areas and each other. According to preliminary results, the 22-year annual profile of NDVI and LAI for AA were between 0.34-0.53 and 0.50-1.07, respectively. Moreover, the NDVI and LAI values vary from 0.26 to 0.63 and 0.30 to 1.48, respectively, between March and October in each year similar to the values found in the literature for the vegetation period for vineyards. The highest NDVI and LAI values were observed in NIA1 and NIA2. The lowest NDVI and LAI values were observed in IA2 and IA3. R (Spearman's correlation coefficients) values for AA are higher between AIA and IA1-IA3 compared to NIA1-NIA2. The lower R values are observed between NIA1 and all other areas.

The trend, seasonal, and residual components in the NDVI and LAI time series were decomposed with seasonal and trend decomposition using LOESS. Moreover, the observations of NDVI and LAI changes were evaluated with meteorological and GPP-related parameters at phenological stages taking into consideration the pre-and post-GPP installations. The results demonstrate the patterns and changes for vineyards based on seasonality and other reasons, and give insight on whether there are any impacts by GPPs.

Keywords: NDVI, LAI, Vineyards, Geothermal Power Plants

References

Food and Agriculture Organization of the United Nations (FAO) (2023). Earth Observation. https://www.fao.org/home/en/

International Renewable Energy Agency (IRENA) (2023). Renewable Capacity Statistics 2023. https://www.irena.org

How to cite: Aydin, M., Leuchner, M., and Kaynak, B.: Observations of Changes in Vineyards Using Long-Term MODIS NDVI and LAI around Geothermal Power Plants, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11264, https://doi.org/10.5194/egusphere-egu24-11264, 2024.

OSPP judging

X4.176

EGU24-14501

Earthquake surface ruptures mornitoring based on deep learning and remote sensing

Yeonju Choi

Surface expression of seismic rupture is a distinctive feature of large earthquakes, and the length of the surface rupture varies from several to several hundred kilometers depending on the magnitude of the earthquake. The structure and mechanical properties of fault zones strongly influence the behavior of earthquake ruptures, and detailed mapping and documentation of fault geometry such as fault bends, steps, branches, and their related segment geometry play a crucial role in determining the propagation and path of earthquake ruptures. The most precise method to map earthquake surface ruptures in detail might be manual mapping in the field by an expert; however, it takes a long time to analyze a vast area, and not everyone has access to the necessary expertise. As an alternative, rupture mapping based on remote sensing has been proposed to supplement the limitations of these field surveys.

In this study, we propose a robust model for automatic detection and analysis of morphological features of surface ruptures using deep learning based on previous research results. Additionally, a skeletonization technique was developed for morphological analysis of detected fractures, and fractures of complex structures on the ground were objectified as individual fractures. Finally, the geometric quantitative characteristics of individualized fractures, including crack location, length, width, and direction, were automatically extracted. By comparing the detection results of the proposed model with the ground truth confirmed by the expert using a line map, the reliability of the entire model could be confirmed.

To examine the applicability of the proposed model, the detection performance of various surface ruptures in localized areas was evaluated, and key characteristics, including rupture direction and pattern for extensive surface ruptures, were clearly identified. The satellite image-based surface destruction detection model proposed in this study can be used as a useful tool for field investigation and earthquake-related basic data collection by automatically detecting various surface destruction and deformations caused by earthquakes. Therefore, the proposed model, which enables fast and accurate fracture and fault mapping and quantitative analysis using high-resolution satellite data, is expected to be utilized as an important integrated solution. It would also be extremely beneficial in characterizing and comprehending the structural, geometrical complexity, and mechanical properties of fault zones.

How to cite: Choi, Y.: Earthquake surface ruptures mornitoring based on deep learning and remote sensing, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14501, https://doi.org/10.5194/egusphere-egu24-14501, 2024.

X4.177

EGU24-19993

The application of UAV Digital Photogrammetry for the investigation and mapping of slow-moving landslides

(withdrawn)

Stefano Devoto

Posters virtual: Wed, 17 Apr, 14:00–15:45 | vHall X4

Display time: Wed, 17 Apr 08:30–Wed, 17 Apr 18:00

Chairpersons: Annalisa Cappello, Gaetana Ganci, Gabor Kereszturi

vX4.33

EGU24-8553

ECS

Mapping Mediterranean temporary ponds in SW Portugal using remote sensing

Ginevra Gennari, Maria C Neves, and Jose P Monteiro

Mediterranean temporary ponds on the Southwest Portuguese coast are areas of significant biodiversity and are recognized as priority habitats by the Habitat Directive (92/43/EEC). The fauna and flora species that inhabit these ponds are well adapted to the variability of the Mediterranean climate, exhibiting a high resistance to drought and the ability to survive without water for months. Among them, the Triops vicentinus species stands out for its uniqueness, considered a living fossil that has persisted since the time of the dinosaurs.

These temporary ponds are situated in shallow depressions where rainwater accumulates for seasonal periods. They face increasing threats due to anthropogenic pressures, primarily arising from intensive farming techniques, real estate development linked to tourism, and a general lack of awareness among the population regarding the significant biological and ecological value of this habitat. The area was the focus of the LIFE + project titled 'Conservation of Temporary Ponds in the Southwest Portuguese Coast of Portugal' conducted between 2013 and 2018. During this initiative, georeferenced mapping of the ponds and associated biodiversity was produced. However, the geographical evolution of these ponds has not been monitored since then.

This study aims to map the ponds and track their evolution between 2018 and the present. To map the water bodies, we use Sentinel-1 and 2 datasets along with the Sentinel application platform SNAP. The surface water extraction method relies on the Normalized Difference Water Index and a supervised classification algorithm. The applied methodology has proven to be efficient in detecting and interpreting the dynamics of water bodies. The results of this investigation enhance our understanding of the uncertainties associated with the applications of Sentinel-2 and Sentinel-1 for monitoring temporary ponds and contribute to our knowledge of the current status of this habitat.

This work was funded by the Portuguese Fundação para a Ciência e a Tecnologia (FCT) I.P./MCTES through national funds (PIDDAC) –

UIDB/50019/2020 (https://doi.org/10.54499/UIDB/50019/2020),

UIDP/50019/2020 (https://doi.org/10.54499/UIDP/50019/2020) and

LA/P/0068/2020(https://doi.org/10.54499/LA/P/0068/2020).

How to cite: Gennari, G., Neves, M. C., and Monteiro, J. P.: Mapping Mediterranean temporary ponds in SW Portugal using remote sensing, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8553, https://doi.org/10.5194/egusphere-egu24-8553, 2024.

vX4.34

EGU24-15824

ECS

Benggang erosion: Critical condition and risk prediction in small watershed

haidong ou, zaijian yuan, and xiankun yang

Benggang is an erosion landform developed on the deep weathered crust in the low mountain and hilly areas of southern China, characterized by wide distribution and high erosion intensity. Similar erosional landforms globally include the lavakas landform in Madagascar, vocorocas in Brazil, and the "collapse" landform in Japan. However, benggang exhibits distinctive development characteristics that differ from these landforms, thus being considered a geomorphic feature unique to southern China. Investigations into benggang reveal that there are approximately 239000 benggangs in southern China, with a total area of 1220 km². Since 1949, benggang erosion has caused the destruction of 360,000 hm² of arable land, damage to over 500,000 houses, and siltation in around 9,000 reservoirs. This has severely hindered residents' lives and economic development, earning it the moniker "ecological ulcer." Previous studies on benggang have primarily focused on the scale of land blocks or sample plots, identifying elevation, slope, and aspect as key factors influencing benggang erosion. Regional-scale studies have found that geological and pedogenic factors, prolonged and intense rainfall, and vegetation cover significantly impact benggang development. However, research on the critical conditions for benggang formation at the small watershed scale is limited. Therefore, this study focuses on the Yuankengshui watershed in Wuhua County, Guangdong Province. Based on drone photogrammetry technology, digital orthophoto images and digital surface models were obtained, and the boundaries of 296 benggangs distributed in the small watersheds were delineated, including the collapsed walls, alluvial bodies, channels and other parts of each benggang. Then, machine learning and multiple linear regression methods are used to analyze the impact of factors such as altitude, slope, aspect, temperature, rainfall, and vegetation cover on the probability of benggang distribution. Results show that benggang is mainly distributed in the 210-230m altitude range, 23-26 ° slope range, 20-60m slope length range, and most of them are located in the sunny slope. Rainstorm is the main reason for the intensified erosion of benggang. Based on the above research, we have established an information based benggang risk prediction model to explore potential benggang occurrence areas and provide reference for future benggang erosion prevention and control.

How to cite: ou, H., yuan, Z., and yang, X.: Benggang erosion: Critical condition and risk prediction in small watershed, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15824, https://doi.org/10.5194/egusphere-egu24-15824, 2024.