How to cite: Teo, H. C., Tan, N. H. L., Zheng, Q., Lim, A. J. Y., Sreekar, R., Chen, X., Zhou, Y., Sarira, T. V., De Alban, J. D. T., Tang, H., Friess, D. A., and Koh, L. P.: Uncertainties in deforestation emission baseline methodologies and implications for carbon markets, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1634, https://doi.org/10.5194/egusphere-egu24-1634, 2024.

The Cerrado-Amazon Transition (CAT) is the world’s largest tropical ecotone and separates the Cerrado Savannah from the Amazon Rainforest. Deforestation and degradation of large swathes of the dense Amazon rainforest and Brazilian Savanna is leading to irreversible transformation and a critical loss of biodiversity. An increase in wildfire and agriculture-led deforestation makes the CAT a dynamic ecological border within the internationally known ‘Arc of Deforestation’. Yet, our understanding of the impacts of deforestation and degradation in the CAT is hampered by a lack of knowledge as to where and when these disturbances occur. Here we combine time-series segmentation and deep learning algorithms to identify and quantify disturbances in the CAT over a 35 - year period. Using a combination of the Landtrendr algorithm, Landsat time series data and a Residual Neural Network (ResNet), we identified four different forest disturbance types (forest clearance, savannah clearance, forest wildfire, savannah wildfire) occurring within the CAT, based on their temporal spectral trajectories. Using our approach, we identified  more than 384,000 km² of disturbance between 1985 and 2020, with forest clearance accounting for the most significant proportion (35%) of the identified change. The accuracy of disturbance detection ranged from 88to 93%, while the accuracy of disturbance type classification reached ~ 95%, although disturbance events occurring within savannas are more difficult to identify, often due to lower initial vegetation cover. The greatest period of disturbance occurred between 1995-1998, due to increased agricultural activity.

How to cite: Li, C., Harris, A., da Conceição Bispo, P., and Dennis, M.: Patterns of disturbance dynamics within the Cerrado-Amazon Transition using time series data and Residual Neural Networks, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2657, https://doi.org/10.5194/egusphere-egu24-2657, 2024.

<p>The global climate situation becomes more and more critical due to the impacts of climate change especially when dealing with flood hazard causing major human and economic losses every year. In Morocco, the Souss-watershed is one of the most vulnerable regions in term of flooding and land degradation. The climate conditions, population growth affect more the land use conditions. The present work introduces a novel approach to assess flood risk in Souss watershed using 4 neural network based models in Google Colab: Artificial neural networks (ANN), Recurrent Neural networks (RNN), One dimensional (1DCNN) and Two dimensional Convolutional neural networks (2DCNN). The models input were constructed using 4 features chosen from 17 of the most triggering flood factors that describe the characteristics of the watershed, including topography, vegetation and soil ones. The Pearsons correlation factor was applied to evaluate the correlation between the features, the variance inflation factor analysis (VIF) was applied to diagnose the collinearity and the Shapley Additive Explanations (SHAP) was applied to evaluate the importance of a factor in the prediction model. For the evaluation and validation process, the calculation of the Mean Absolute Error (MAE) and loss was used to evaluate the accuracy of predictions along with the calculation of the ROC (Receiver Operating Characteristic) and the AUC (Area Under the ROC Curve) to compare between the four models, the results demonstrated that the RNN has the highest performance with an accuracy of 96% and a validation loss of 0.0984 and a validation MAE of 0.2553, followed by the ANN with a slightly lower accuracy of 95% , 2DCNN and 1DCNN demonstrated lower accuracies of 87% and 81%. These findings have demonstrated that in the flood susceptibility mapping context, the application of complex neural networks such as 1DCNN and 2DCNN calls for more tuning and optimizing to overcome over-fitting issues, and that using simple neural networks such as RNN and ANN can be more effective in achieving more accurate predictions.</p>

How to cite: Aghenda, M., Labbaci, A., Hssaisoune, M., and Bouchaou, L.: Flood susceptibility mapping using neural network based models in Morocco: Case of Souss Watershed, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3447, https://doi.org/10.5194/egusphere-egu24-3447, 2024.

The cultivation of vine is an important economic sector in agriculture as well as a cultural legacy
in many regions. Lower Franconia is one of Germany’s largest wine-producing areas with more
than 6,000 ha of vineyards and a production of 450,000 hectolitres of wine. In the context of climate
change, strategies to adapt to changing precipitation and temperature patterns and to mitigate risks
from drought and grape diseases is of utter importance for sustainable viticulture. Water deficit
produces diverse effects, such as reduced berry size or the failure of fruit maturation, depending on the
plant’s growth stage [1]. Severe water stress triggers partial or complete stomatal closure, resulting in a
reduction of photosynthetic activity [3]. Thus, ideal soil moisture is key to sustainable viticulture and
monitoring plant development, soil moisture and climate variables is crucial for precision viticulture
[2]. Due to the typical trellis systems, satellite remote sensing using deca-meter resolutions (such
as Landsat or Sentinel-2 series) is not well-suited for plant monitoring as pixel information at this
resolution consists of vines and ground information of the space between plant rows that might be
grass, other plants or soil. Thus, we investigate time-series of very-high-resolution multispectral
data derived from a UAV-based system, hyperspectral data from in-situ measurements as well as
sensor data for soil moisture and evaluate results with respect to different irrigation patterns. Such
multi-sensor and multi-temporal approaches contribute to a better understanding of the vineyard
as a dynamic system and, thus, lead to better monitoring and management options and allow to
build a GIS-based Digital Twin of the vineyard. We will present first results from daily to weekly
measurements, evaluate different vegetation indices and highlight temporal patterns and reaction
times between soil moisture data and spectral measurements.

[1] WJ Hardie and JA Considine. “Response of grapes to water-deficit stress in particular stages of
development”. In: American Journal of Enology and Viticulture 27.2 (1976), pp. 55–61.
[2] Margareth A Oliver. “An overview of geostatistics and precision agriculture”. In: Geostatistical
applications for precision agriculture (2010), pp. 1–34.
[3] Maria Romero et al. “Vineyard water status estimation using multispectral imagery from an UAV
platform and machine learning algorithms for irrigation scheduling management”. In: Computers
and Electronics in Agriculture 147 (Apr. 2018), pp. 109–117. (Visited on 10/10/2023).

How to cite: Brandmeier, M., Heßdörfer, D., Meyer-Spelbrink, A., Siebenlist, P., and Kraus, A.: Time-series analysis on multi-modal data for precisionviticulture to support adaptation to changing climate patternsin Franconia, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4541, https://doi.org/10.5194/egusphere-egu24-4541, 2024.

Traditional imageery workflows in many cases still follow a Datset --> process --> result Dataset scheme.
In todays world of high spatial and temporal imagery resolution and analytical cadence, this turns out to
be increasingly inpractical. This is backed by todays cloud based mass data storage and constant change detection requirements.
So accessing Imagery as dynamic compilations and analyzing stacks, time-series, blocks upon demand, largely without producing
permanent internmeidate data has become a paradigm. Added to it is the demand to analyze the Imagery sources together with in-situ
sources, sensors, base geometries and other georelated data sources.

This presentations will outline the capabilities and strategies of the ArcGIS Platform to fulfill these demands and reference
cloud based data and processing as well as dynamic server analytics and traditional Desktop approaches - UI and Python/Notebook driven.
Time-series analysis, DeepLearning/AI and combined Raster/Vector analysis will be used as examples. 

Target audience is anyone interested in geospatial analysis combining Imagery and other geospatial data sources.

How to cite: Doerffel, G.: Image Management and Analytics integrated in geospatial workflows, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6098, https://doi.org/10.5194/egusphere-egu24-6098, 2024.

Perennial crops hold significant importance in global agricultural markets, local agricultural economies, poverty reduction, and biodiversity conservation. Accurate spatial distribution data of different perennial crops are crucial for local agricultural management, crop yield prediction, and sustainable agricultural development. However, precisely obtaining the spatial distribution of different perennial crops using remote sensing data, especially in regions with complex cropping patterns, remains a challenge. This study, focusing on Yantai City, China, and three counties in California, USA, aims to develop a method suitable for both smallholder and intensive production systems. On the Google Earth Engine (GEE) platform, we applied linear spectral mixture analysis (LSMA) to transform Sentinel-2 time-series data (2020-2022) from the original spectral space into a unified endmember space, including photosynthetic vegetation, non-photosynthetic vegetation, soil, and shadow, thereby characterizing the time-series land surface component information. Subsequently, based on the time-series endmembers data, we quantified seven sets of harmonic features and five phenological features. Relative phases were employed as phase indicators for the harmonic features. These seven harmonic features represent the intra-annual patterns of land surface component changes and inter-component interactions, while the phenological features quantify the timing of phenological events. Our findings reveal significant variations in the intra-annual patterns of land surface component changes among different perennial crops, attributable to differences in phenology and phenology-associated human management. Building on this, the study achieved precise mapping of different perennial crops, even in areas with complex cropping patterns. The overall classification accuracy for perennial crops in Yantai City and the three counties in California was 90.3% and 94.8%, respectively, with Kappa coefficients of 89.2% and 93.9%. Utilizing intra-annual time-series land surface component information for extracting the spatial distribution of perennial crops demonstrated advantages over traditional optical indices. This work provides a method that is applicable to both smallholder and intensive production systems, enabling precise mapping of perennial crop types. It represents an important step towards achieving large-scale mapping of perennial crop types.

How to cite: Gao, X., Hu, Q., Lun, F., and Sun, D.: Improved Mapping of Perennial Crop Types Based on Patterns of Intra-Annual Variation in Land Surface Components, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10788, https://doi.org/10.5194/egusphere-egu24-10788, 2024.

Weather extremes severely affect agricultural production and threaten food security worldwide. Throughout the growing season, crops can experience various degrees of weather stress whereby multiple stressors could occur simultaneously or intermittently. For large spatial extents, it is difficult to estimate actual crop damage merely through field experiments or crop models. Remote sensing can help to detect crop damage and estimate lost yield due to weather extremes over large spatial extents, but current RS-based studies usually focus on a single stress or event. We propose a novel scalable method to predict in-season yield losses at the sub-field level and attribute these losses to different weather extremes. To assess our method’s potential, we conducted a proof-of-concept case study on winter cereal paddocks in South Australia using data from 2017 to 2022. To detect crop growth anomalies throughout the growing season, we aligned a two-band Enhanced Vegetation Index (EVI2) time series from Sentinel-2 with thermal time derived from gridded meteorological data. The deviation between the expected and observed EVI2 time series was defined as the Crop Damage Index (CDI). We assessed the performance of the CDI within specific phenological windows to predict yield loss. Finally, by comparing instances of substantial increase in CDI with different extreme weather indicators, we explored which (combinations of) extreme weather events were likely responsible for the experienced yield reduction. We found that the use of thermal time diminished the temporal deviation of EVI2 time series between years, resulting in the effective construction of typical stress-free crop growth curves. Thermal-time-based EVI2 time series resulted in better prediction of yield reduction than those based on calendar dates. Yield reduction could be predicted before grain-filling (approximately two months before harvest) with an R² of 0.83 for wheat and 0.91 for barley. The combined analysis of CDI curves and extreme weather indices allowed for timely detection of weather-related causes of crop damage, which also captured the spatial variations of crop damage attribution at sub-field level. Our approach can help to improve early assessment of crop damage and understand weather causes of such damage, thus informing strategies for crop protection.

How to cite: Duan, K., Vrieling, A., Schlund, M., Nidumolu, U. B., Ratcliff, C., Collings, S., and Nelson, A.: Sub-field detection of cereal yield losses and its causes using Sentinel-2 time series and weather data, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10831, https://doi.org/10.5194/egusphere-egu24-10831, 2024.

Pastures and grasslands are the largest land cover of Earth's surface, comprising fundamental landscapes for water and nutrient cycling, food production, biodiversity conservation and land management in the planet. Monitoring the conditions and productivity aspects of these lands can lead to major contributions for land degradation mitigating in line with sustainable development goals defined by the United Nations (UN) 2030 agenda. Nevertheless, an operational approach able to monitor productivity of pastures and grasslands at global scale and high spatial resolution (e.g. 30-m) is a challenging research problem. Aiming to contribute with this topic, the current work present a methodology to derive 30-m bi-monthly time-series of Gross Primary Productivity (GPP) for pastures and grasslands of the world based on GLAD Landsat ARD (collection-2) and a customized Light Use Efficiency Model (LUE). The Landsat imagery were aggregated by every two months and gapfilled by an temporal interpolation based on Fast Fourier Transform (FFT). The complete, consistent and gapfilled Landsat time-series was used to estimate the Fraction of Photosynthetically Active Radiation (FPAR) and Land Surface Water Index (LSWI), which combined with 1-km MODIS temperature (MOD11A2) and 1° CERES Photosynthetically Active Radiation (SYN1deg v4.1 - PAR) images resulted in 30-m global GPP time-series product from 2000 onwards. Our preliminary validation approach, based on FLUXNET2015 data, indicated a R2 of 0.67 and RMSE of 2.06 for in-situ stations located Europe. We are working to release the first version of the product as open data (CC-BY license) in the context of the Global Pasture Watch project and the World Resources Institute's Land & Carbon Lab, establishing partnerships with local organization and research institute to collect feedback and additional validation data to improve further versions of the product.

How to cite: Parente, L., Hackländer, J., Consoli, D., Hengl, T., Mesquita, V., Ferreira, L., Sloat, L., Teles, N., Ho, Y.-F., Wheeler, I., and Stolle, F.: Global pastures and grasslands productivity time series mapped at 30-m spatial resolution using Light Use Efficiency Model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13195, https://doi.org/10.5194/egusphere-egu24-13195, 2024.

Remote sensing satellites provide key data that can be used to better understand the Earth, respond to hazardous events, and to make decisions related to climate change, population growth, and more. For decades, many space agencies have provided high quality remote sensing data free of charge for their end-users. Although these data have been accessible, and widely used, the raw remote sensing measurements can be challenging to analyze for non-specialists. Furthermore, the large quantity of data available today makes it nearly impossible to perform large scale analysis on personal computers. To overcome these barriers, the Observational Products for End-Users from Remote Sensing Analysis (OPERA) project, led by the Jet Propulsion Laboratory, with project partners from NASA, USGS, and academia, are producing three analysis ready data products derived from satellite-based synthetic aperture radar (SAR) and optical data. These products were designed to meet the needs of U.S. federal agencies that were identified by the Satellite Needs Working Group (an initiative of the U.S. Group on Earth Observations). The OPERA analysis ready data products are derived from the NASA/USGS Landsat 8/9 sensors, ESA’s Sentinel-1 and -2, and NASA-ISRO SAR Mission (NISAR). Specific products include: (1) a near-global Dynamic Surface Water eXtent (DSWx) product suite from optical and SAR data, (2) a near-global Surface Disturbance (DIST) product suite from optical and SAR data, and (3) a North America Displacement (DISP) product suite from SAR data. In addition, OPERA is producing intermediate-level products including: (1) a North America Land Coregistered Single-Look Complex (CSLC) product from SAR data, and (2) a near-global land surface Radiometric Terrain Corrected (RTC) SAR backscatter product from SAR data. These two intermediate SAR products allow for user-customized product generation. All OPERA products are freely available and all OPERA software are open-access (https://github.com/opera-adt). More information on OPERA can be found at https://www.jpl.nasa.gov/go/opera. 

Here, we present the latest OPERA project updates. We provide an overview of OPERA’s in-production products, which include DSWx, DIST, RTC, and CSLC, and information about OPERA’s future DISP product. We will showcase product use-cases, with a focus on detection and monitoring of hazards such as floods, wildfires, earthquakes, and landslides. We also discuss how the free and open OPERA data can be accessed through the NASA Distributed Active Archive Centers. Finally, we demonstrate our open OPERA application tools (https://github.com/OPERA-Cal-Val/OPERA_Applications) that are designed to increase data use and discoverability.

How to cite: Handwerger, A. L., Chan, S., and Bekaert, D. and the OPERA Team: The Observational Products for End-Users from Remote Sensing Analysis (OPERA) Project: Status, Access, Applications, Tools, and More, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14221, https://doi.org/10.5194/egusphere-egu24-14221, 2024.

This talk will discuss techniques used to map aboveground biomass in the region of Cantabria, Spain, and examine tree cover change over time. Using the biomass workflow outlined by Esri's Hong Xu we leverage machine learning tools in ArcGIS Pro, altimeter sensor data, optical satellite imagery, and DEM data to gain foundational insights into land cover and forest density. By combining trajectory point data from GEDI, which contains aboveground biomass information, with surface reflectance bands from Landsat collection 2, and DEM data we examine forest density, and coverage in the Cantabria region. These results are compared to data analyzed from previous decades using deep learning techniques to assess forest change in the region. Using ArcGIS Pro and AWS we can integrate remote sensing datasets from multiple satellite sensors and utilize machine learning tools to train and run a regression model providing us with an estimation of the above-ground biomass for the region of Cantabria, Spain in the form of a raster layer. This data can then be used in conjunction with results from tree cover analysis over time using deep learning and displayed in an interactive web application. This multifaceted analysis can provide researchers, policymakers, and stakeholders with key insights into progress and prioritization and aid in addressing local challenges related to forest health. 

How to cite: Graham, L., Alouta, R., Tanh, L., and Xu, H.: Examining aboveground biomass and tree cover change in the region of Cantabria, Spain using multiple satellite sensors and machine learning tools in ArcGIS Pro. , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19566, https://doi.org/10.5194/egusphere-egu24-19566, 2024.

Land surface temperature (LST) is of fundamental importance to many aspects of geosciences, such as, net radiation budget, evaluation and monitoring of forest and crops. The LST is also a key-parameter to derive the Surface Urban Heat Island (SUHI) or health related indices, for example, the Discomfort Index (DI). It is therefore an essential prerequisite to understand the urban climate and to support the definition of mitigation strategies, health risk management plans, public policies among other initiatives to effectively address the adverse effects of heat. While the LST is commonly retrieved from data acquired in the TIR (Thermal InfraRed) spectral domain by remote multispectral sensors with about 1K accuracy for natural surfaces, its retrieval over urban areas is not trivial. Urban landscapes possess tremendous challenges in LST estimates due to its high heterogeneity of surfaces and materials, and the three-dimensional (3D) configuration of the elements that are present on urban areas.

The Thermal InfraRed Imaging Satellite for High-resolution Natural resource Assessment (TRISHNA) is planned for launch in 2025 and features a TIR instrument that will image the Earth every three days, at 57 m resolution, providing the research community with critical information to understand the radiative interactions and impacts al local level. This new satellite mission will offer an unprecedented opportunity to support urban microclimate studies. Based on extensive radiative transfer simulations using the Discrete Anisotropic Radiative Transfer Model (DART) and sensitivity analysis, this work investigates the impacts of 3D urban structures (e.g., road width, building height, building density) and materials with different optical properties on LST estimation at the TRISHNA spatial resolution. In fine, the idea is to develop a method to minimize these impacts on LST estimated from the TRISHNA data.

First, a processing chain has been set up to simulate TRISHNA LST with DART, by using as inputs i) the configuration of the sensor and ii) 3D urban forms with different geometric and optical properties. Second, the radiative transfer modeling for simulation of the TIR remote sensing signal is performed. Finally, by correlating the simulated TRISHNA LST and the surface characteristics for each scene, the main parameters impacting the LST in urban environments have been identified. From these results, a correction method at satellite scale to minimize the impacts of urban 3D variables on LST will be formulated.

How to cite: Cifuentes La Mura, R., Roupioz, L., Nerry, F., and Briottet, X.: Investigating the impact of urban 3D variables on satellite land surface temperature estimates, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20017, https://doi.org/10.5194/egusphere-egu24-20017, 2024.

Consistent and continuous Land Surface Temperature at high temporal resolution is essential for many applications, such as anomaly detection (e.g. agricultural droughts), urban heat island monitoring or irrigation and crop water stress, among others. LST can be retrieved at high spatial resolution from spaceborne thermal infrared (TIR) instruments, like MODIS/VIIRS, Landsat, ECOSTRESS, and ASTER. But these data come with large temporal gaps due to cloud cover and orbit/sensor characteristics and consequently complicate time series analyses. 

To overcome these limitations, we developed a daily 100m LST product based on the synergy between passive microwave brightness temperatures from the Advanced Microwave Scanning Radiometer 2 (AMSR2), and optical data from Sentinel 2 within a novel disaggregation method [1]. This results in a dataset that allows for monitoring environmental systems consistently and continuously in near-real time. The passive microwave observations offer a distinct advantage in LST estimation due to the ability to penetrate cloud cover and measure thermal emissions at the surface. On the other hand, Sentinel-2, with its high spatial resolution multispectral bands, provides rich information on land cover and land surface properties. By combining these complementary datasets, we aim to leverage the strengths of both sensors to improve the accuracy and spatial resolution of LST retrieval. The method uses the abundance of overlaps between passive microwave footprints in combination with higher spatial information from S2 NIR and SWIR for downscaling at 100m resolution since 2017 at 1:30am and at 1:30pm. 

To assess the accuracy of the 100m LST, we compared the time series of microwave-based LST at 100+ locations against in situ measurements, MODIS and Landsat LST data. Currently, the temporal accuracy compared to these in-situ stations is ±3.1K with a correlation of 0.91 (for MODIS this was 2.6 and 0.94). We also performed a spatial comparison of our 100m LST data over agricultural regions against Landsat LST. While the few clear-sky Landsat LST observations is a limitation for the comparison, the preliminary results show a spatial accuracy between ±1.5K and ±4K.

Our results demonstrate that our LST data-fusion approach is a viable methodology to generate a temporally and spatially high resolution LST archive. We are aiming to bridge current and future missions for high-resolution LST by harnessing the complementary capabilities of multi-sensor data fusion. The proposed framework holds great potential for improving our understanding and monitoring Earth's complex environmental systems, such as local surface energy dynamics, climate processes or supporting various environmental applications requiring accurate and high-resolution LST information.

[1] de Jeu, R. A. M., de Nijs, A. H. A., & van Klink, M. H. W. (2017). Method and system for improving the resolution of sensor data. https://patents.google.com/patent/WO2017216186A1/en 

How to cite: Dijkstra, J., Malbeteau, Y., Ghironi, M., Guillevic, P., and de Jeu, R.: Unlocking time series analysis with data-fusion land surface temperature , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20629, https://doi.org/10.5194/egusphere-egu24-20629, 2024.