The species richness and biodiversity of vegetation in Hungary are increasingly threatened by invasive plant species brought in from other continents and foreign ecosystems. These invasive plant species have spread aggressively in the natural and semi-natural habitats of Europe. Common milkweed (Asclepias syriaca) is one of the species that pose the greatest ecological menace. Therefore, the primary purpose of the present study is to map and monitor the spread of common milkweed, the most common invasive plant species in Europe. Furthermore, the possibilities to detect and validate this special invasive plant by analyzing hyperspectral remote sensing data were investigated. In combination with field reference data, high-resolution hyperspectral aerial images acquired by an unmanned aerial vehicle (UAV) platform in 138 spectral bands in areas infected by common milkweed were examined. Then, support vector machine (SVM) and artificial neural network (ANN) classification algorithms were applied to the highly accurate field reference data. As a result, common milkweed individuals were distinguished in hyperspectral images, achieving an overall accuracy of 92.95% in the case of supervised SVM classification. Using the ANN model, an overall accuracy of 99.61% was achieved. To evaluate the proposed approach, two experimental tests were conducted, and in both cases, we managed to distinguish the individual specimens within the large variety of spreading invasive species in a study area of 2 ha, based on centimeter spatial resolution hyperspectral UAV imagery.

Today, invasive alien species cause serious trouble for biodiversity and ecosystem services, which are essential for human survival. In order to effectively manage invasive species, it is important to know their current distribution and the dynamics of their spread. Unmanned aerial vehicle (UAV) monitoring is one of the best tools for gathering this information from large areas. Vegetation indices for multispectral camera images are often used for this, but RGB colour-based vegetation indices can provide a simpler and less expensive solution. The goal was to examine whether six RGB indices are suitable for identifying invasive plant species in the QGIS environment on UAV images. To examine this, we determined the shoot area and number of common milkweed (Asclepias syriaca) and the inflorescence area and number of blanket flowers (Gaillardia pulchella) as two typical invasive species in open sandy grasslands. According to the results, the cover area of common milkweed was best identified with the TGI and SSI indices. The producers’ accuracy was 76.38% (TGI) and 67.02% (SSI), while the user’s accuracy was 75.42% (TGI) and 75.12% (SSI), respectively. For the cover area of blanket flower, the IF index proved to be the most suitable index. In spite of this, it gave a low producer’s accuracy of 43.74% and user’s accuracy of 51.4%. The used methods were not suitable for the determination of milkweed shoot and the blanket flower inflorescence number, due to significant overestimation. With the methods presented here, the data of large populations of invasive species can be processed in a simple, fast, and cost-effective manner, which can ensure the precise planning of treatments for nature conservation practitioners.

Classification of multispectral optical satellite data using machine learning techniques to derive land use/land cover thematic data is important for many applications. Comparing the latest algorithms, our research aims to determine the best option to classify land use/land cover with special focus on temporary inundated land in a flat area in the south of Hungary. These inundations disrupt agricultural practices and can cause large financial loss. Sentinel 2 data with a high temporal and medium spatial resolution is classified using open source implementations of a random forest, support vector machine and an artificial neural network. Each classification model is applied to the same data set and the results are compared qualitatively and quantitatively. The accuracy of the results is high for all methods and does not show large overall differences. A quantitative spatial comparison demonstrates that the neural network gives the best results, but that all models are strongly influenced by atmospheric disturbances in the image.

Monitoring and quantifying the development of drought extremes is important to agriculture, water, and land management. For this, soil moisture (SM) is an effective indicator. However, currently, real-time monitoring and forecasting of SM is challenging. Thus, this study develops and tests a methodology based on machine learning methods that integrates ground-based data, Sentinel-1 satellite soil moisture (S1SSM) data, meteorological data, and relevant environmental parameters to improve the estimation of the spatiotemporal changes in SM. It also evaluates the relevance of the applied parameters and the applicability and limitations of S1SSM data in SM monitoring. Specifically, the performances of four machine learning methods (multiple linear regression, support vector machine regression, extreme gradient boosting, and a deep neural network) were evaluated in an area increasingly exposed to hydrological extremes. Overall, the extreme gradient boosting model provided the best result (R² = 0.92). In this case, the difference between the modeled and observed SM values at ground-based stations was below 3%, with only five stations reporting differences above 5%, indicating the effectiveness of this model for SM monitoring in larger areas. Additionally, the spatial pattern of the observed S1SSM values and the modeled values showed good agreement (with a difference below 10%) in the larger part (45.5%) of the area, while more than 20% difference occurred in 27.1% of the area, demonstrating the application potential of S1SSM data in areas with less heterogeneous land use. However, the results also suggest that the S1SSM data can be affected by land use and/or soil types.

Changing climate is expected to cause more extreme weather patterns in many parts of the world. In the Carpathian Basin, it is expected that the frequency of intensive precipitation will increase causing inland excess water (IEW) in parts of the plains more frequently, while currently the phenomenon already causes great damage. This research presents and validates a new methodology to determine the extent of these floods using a combination of passive and active remote sensing data. The method can be used to monitor IEW over large areas in a fully automated way based on freely available Sentinel-1 and Sentinel-2 remote sensing imagery. The method is validated for two IEW periods in 2016 and 2018 using high-resolution optical satellite data and aerial photographs. Compared to earlier remote sensing data-based methods, our method can be applied under unfavorite weather conditions, does not need human interaction and gives accurate results for inundations larger than 1000 m². The overall accuracy of the classification exceeds 99%; however, smaller IEW patches are underestimated due to the spatial resolution of the input data. Knowledge on the location and duration of the inundations helps to take operational measures against the water but is also required to determine the possibilities for storage of water for dry periods. The frequent monitoring of the floods supports sustainable water management in the area better than the methods currently employed.

Floodplains are a fundamental source of multiple functions and services. Despite their various benefits, a dramatic reduction in floodplain areas has occurred in most large river systems over the last few centuries, and is still ongoing. Human modifications (such as river regulation, dam construction, and land use changes) due to economic growth, increasing population size, etc., were and still are drivers of major floodplain losses. Therefore, studies offering solutions for floodplain preservation and restoration are of great importance for sustainable floodplain management. This paper presents methods to identify active, former, and potential floodplains, and their application to the Danube River. We used hydraulic data, historical sources, and recent geospatial data to delineate the three floodplain types. Fifty hydraulically active floodplains larger than 500 ha were identified. According to our results, the extent of Danube floodplains has been reduced by around 79%. With the support of different representatives from the Danube countries, we identified 24 potential floodplains. However, the share of active and potential floodplains in relation to former floodplains ranges between 5% and 49%, demonstrating the huge potential for additional restoration sites. This analysis contributes to an understanding of the current and the past floodplain situation, increases awareness of the dramatic floodplain loss along the Danube, and serves as a basis for future floodplain management.