Search

The quantification of emissions of greenhouse gases and air pollutants through the inversion of plumes in satellite images remains a complex problem that current methods can only assess with significant uncertainties. The anticipated launch of the CO2M (Copernicus Anthropogenic Carbon Dioxide Monitoring) satellite constellation in 2026 is expected to provide high-resolution images of CO2 (carbon dioxide) column-averaged mole fractions (XCO2), opening up new possibilities. However, the inversion of future CO2 plumes from CO2M will encounter various obstacles. A challenge is the low CO2 plume signal-to-noise ratio due to the variability in the background and instrumental errors in satellite measurements. Moreover, uncertainties in the transport and dispersion processes further complicate the inversion task. To address these challenges, deep learning techniques, such as neural networks, offer promising solutions for retrieving emissions from plumes in XCO2 images. Deep learning models can be trained to identify emissions from plume dynamics simulated using a transport model. It then becomes possible to extract relevant information from new plumes and predict their emissions. In this paper, we develop a strategy employing convolutional neural networks (CNNs) to estimate the emission fluxes from a plume in a pseudo- XCO2 image. Our dataset used to train and test such methods includes pseudo-images based on simulations of hourly XCO2 , NO2 (nitrogen dioxide), and wind fields near various power plants in eastern Germany, tracing plumes from anthropogenic and biogenic sources. CNN models are trained to predict emissions from three power plants that exhibit diverse characteristics. The power plants used to assess the deep learning model's performance are not used to train the model. We find that the CNN model outperforms state-of-the-art plume inversion approaches, achieving highly accurate results with an absolute error about half of that of the cross-sectional flux method and an absolute relative error of ∼ 20 % when only the XCO2 and wind fields are used as inputs. Furthermore, we show that our estimations are only slightly affected by the absence of NO2 fields or a detection mechanism as additional information. Finally, interpretability techniques applied to our models confirm that the CNN automatically learns to identify the XCO2 plume and to assess emissions from the plume concentrations. These promising results suggest a high potential of CNNs in estimating local CO2 emissions from satellite images. [ABSTRACT FROM AUTHOR]

The quantification of emissions of greenhouse gases and air pollutants through the inversion of plumes in satellite images remains a complex problem that current methods can only assess with significant uncertainties. The anticipated launch of the CO2M satellite constellation in 2026 is expected to provide high-resolution images of CO2 column-averaged mole fractions (XCO2), opening up new possibilities. However, the inversion of future CO2 plumes from CO2M will encounter various obstacles. A challenge is the CO2 plume low signal-to-noise ratio, due to the variability of the background and instrumental errors in satellite measurements. Moreover, uncertainties in the transport and dispersion processes further complicate the inversion task. To address these challenges, deep learning techniques, such as neural networks, offer promising solutions for retrieving emissions from plumes in XCO2 images. Deep learning models can be trained to identify emissions from plume dynamics simulated using a transport model. It then becomes possible to extract relevant information from new plumes and predict their emissions. In this paper, we employ convolutional neural networks (CNN) to estimate the emission fluxes from a plume in a pseudo XCO2 image. Our dataset used to train and test such methods includes pseudo images based on simulations of hourly XCO2, NO2 and wind fields near various power plants in Eastern Germany, tracing plumes from anthropogenic and biogenic sources. CNN models are trained to predict emissions from three power plants that exhibit diverse characteristics. The power plants used to assess the deep learning model's performance are not used to train the model. We find that the CNN model outperforms state of the art plume inversion approaches, achieving highly accurate results with an absolute error about half of that of the cross-sectional flux method. Furthermore, we show that our estimations are only slightly affected by the absence of NO2 fields or a detection mechanism as additional information. Finally, interpretability techniques applied to our models confirm that the CNN automatically learns to identify the XCO2 plume and to assess emissions from the plume concentrations. These promising results suggest a high potential of CNNs in estimating local CO2 emissions from satellite images. [ABSTRACT FROM AUTHOR]

The use of deep learning (DL) with convolutional neural networks (CNN) to monitor surface water can be a valuable supplement to costly and labour-intense standard gauging stations. This paper presents the application of a recent CNN semantic segmentation method (SegNet) to automatically segment river water in imagery acquired by RGB sensors. This approach can be used as a new supporting tool because there are only a few studies using DL techniques to monitor water resources. The study area is a medium-scale river (Wesenitz) located in the East of Germany. The captured images reflect different periods of the day over a period of approximately 50 days, allowing for the analysis of the river in different environmental conditions and situations. In the experiments, we evaluated the input image resolutions of 256 × 256 and 512 × 512 pixels to assess their influence on the performance of river segmentation. The performance of the CNN was measured with the pixel accuracy and IoU metrics revealing an accuracy of 98% and 97%, respectively, for both resolutions, indicating that our approach is efficient to segment water in RGB imagery. [ABSTRACT FROM AUTHOR]

We present an approach for detecting early signs for upcoming forest damages by training a Convolutional Neural Network (CNN) for the pixel-wise prediction of the remaining life-time (RLT) of trees in forests based on Sentinel-2 imagery. We focus on a scenario in which reference data are only available for a related task, namely for a bi-temporal pixel-wise classification of forest degradation. This reference is used to train a CNN for the pixel-wise prediction of forest degradation. In this context, we propose a new sub-sampling-based approach for compensating the effects of a heavy class imbalance in the training data. Using the resulting classification model, we predict semi-labels for images of a Sentinel-2 time series, from which training data for a CNN designed to regress the RLT can be derived after some label cleansing. However, due to data gaps in the time series, e.g. caused by clouds, only intervals can be derived for the target variable to be regressed, and for some training pixels one of the interval limits may even be unknown. Consequently, we propose a new loss function for training a CNN for regressing the RLT that only requires the known interval limits. The method is evaluated on a data set in Germany, covering a time-span of 5 years. We show that the proposed sub-sampling strategy for dealing with strong label imbalance when training the classifier significantly reduces the training time compared to other approaches. We further show that our model predicts the RLT with a maximum error of two months for 80% of the forest pixels that die within one year from the acquisition date of the Sentinel-2 image. [ABSTRACT FROM AUTHOR]

The application of machine learning (ML) including deep learning models in hydrogeology to model and predict groundwater level in monitoring wells has gained some traction in recent years. Currently, the dominant model class is the so-called single-well model, where one model is trained for each well separately. However, recent developments in neighbouring disciplines including hydrology (rainfall–runoff modelling) have shown that global models, being able to incorporate data of several wells, may have advantages. These models are often called "entity-aware models", as they usually rely on static data to differentiate the entities, i.e. groundwater wells in hydrogeology or catchments in surface hydrology. We test two kinds of static information to characterize the groundwater wells in a global, entity-aware deep learning model set-up: first, environmental features that are continuously available and thus theoretically enable spatial generalization (regionalization), and second, time-series features that are derived from the past time series at the respective well. Moreover, we test random integer features as entity information for comparison. We use a published dataset of 108 groundwater wells in Germany, and evaluate the performance of the models in terms of Nash–Sutcliffe efficiency (NSE) in an in-sample and an out-of-sample setting, representing temporal and spatial generalization. Our results show that entity-aware models work well with a mean performance of NSE >0.8 in an in-sample setting, thus being comparable to, or even outperforming, single-well models. However, they do not generalize well spatially in an out-of-sample setting (mean NSE <0.7 , i.e. lower than a global model without entity information). Strikingly, all model variants, regardless of the type of static features used, basically perform equally well both in- and out-of-sample. The conclusion is that the model in fact does not show entity awareness, but uses static features merely as unique identifiers, raising the research question of how to properly establish entity awareness in deep learning models. Potential future avenues lie in bigger datasets, as the relatively small number of wells in the dataset might not be enough to take full advantage of global models. Also, more research is needed to find meaningful static features for ML in hydrogeology. [ABSTRACT FROM AUTHOR]

Our subject is a new catalogue of radar-based heavy rainfall events (CatRaRE) over Germany and how it relates to the concurrent atmospheric circulation. We classify daily ERA5 fields of convective indices according to CatRaRE, using an array of 13 statistical methods, consisting of 4 conventional ("shallow") and 9 more recent deep machine learning (DL) algorithms; the classifiers are then applied to corresponding fields of simulated present and future atmospheres from the Coordinated Regional Climate Downscaling Experiment (CORDEX) project. The inherent uncertainty of the DL results from the stochastic nature of their optimization is addressed by employing an ensemble approach using 20 runs for each network. The shallow random forest method performs best with an equitable threat score (ETS) around 0.52, followed by the DL networks ALL-CNN and ResNet with an ETS near 0.48. Their success can be understood as a result of conceptual simplicity and parametric parsimony, which obviously best fits the relatively simple classification task. It is found that, on summer days, CatRaRE-convective atmospheres over Germany occur with a probability of about 0.5. This probability is projected to increase, regardless of method, both in ERA5-reanalyzed and CORDEX-simulated atmospheres: for the historical period we find a centennial increase of about 0.2 and for the future period one of slightly below 0.1. [ABSTRACT FROM AUTHOR]

Benefiting from the high cadence and spatial resolution of the new generation of Earth observation satellites, remote sensing technology is allowing us to derive more valuable information for the agricultural sector. Crop classification is one of the fundamental information derivatives from Earth observation data researchers used for food security, crop monitoring, and economic assessment. The robustness of a crop classification model to variations in environmental and management conditions due to time and location is one of the crucial requirements. To achieve this, we developed a novel self-supervised method using the advantage of unlabeled samples and transformer architectures. We used six different areas in Germany and four years to evaluate the robustness of the model. Our experiments showed that self-supervised deep learning methods could provide a significant advantage in handling these variations. In some cases, we observed around 30 percentage points improvements in F1-score performance compared to a Random Forest based model. [ABSTRACT FROM AUTHOR]

Tropospheric ozone is a secondary air pollutant that is harmful to living beings and crops. Predicting ozone concentrations at specific locations is thus important to initiate protection measures, i.e. emission reductions or warnings to the population. Ozone levels at specific locations result from emission and sink processes, mixing and chemical transformation along an air parcel's trajectory. Current ozone forecasting systems generally rely on computationally expensive chemistry transport models (CTMs). However, recently several studies have demonstrated the potential of deep learning for this task. While a few of these studies were trained on gridded model data, most efforts focus on forecasting time series from individual measurement locations. In this study, we present a hybrid approach which is based on time-series forecasting (up to 4 d) but uses spatially aggregated meteorological and chemical data from upstream wind sectors to represent some aspects of the chemical history of air parcels arriving at the measurement location. To demonstrate the value of this additional information, we extracted pseudo-observation data for Germany from a CTM to avoid extra complications with irregularly spaced and missing data. However, our method can be extended so that it can be applied to observational time series. Using one upstream sector alone improves the forecasts by 10 % during all 4 d, while the use of three sectors improves the mean squared error (MSE) skill score by 14 % during the first 2 d of the prediction but depends on the upstream wind direction. Our method shows its best performance in the northern half of Germany for the first 2 prediction days. Based on the data's seasonality and simulation period, we shed some light on our models' open challenges with (i) spatial structures in terms of decreasing skill scores from the northern German plain to the mountainous south and (ii) concept drifts related to an unusually cold winter season. Here we expect that the inclusion of explainable artificial intelligence methods could reveal additional insights in future versions of our model. [ABSTRACT FROM AUTHOR]