TY - CONF AU - Rapai, Tibor AU - Baják, Petra AU - Lukács, András AU - Székely, Balázs AU - Erőss, Anita TI - Understanding near-surface hydrogeological processes around Lake Velence (Hungary) – using mesh graph neural networks on multidimensional remote sensing data T2 - EGU General Assembly 2024 : abstracts PB - European Geosciences Union (EGU) C1 - Wien PY - 2024 DO - 10.5194/egusphere-egu24-5561 UR - https://m2.mtmt.hu/api/publication/34802132 ID - 34802132 AB - Lake Velence is a shallow soda lake in Hungary whose water budget is mainly driven by precipitation and evaporation. The lake has shown a deteriorating tendency recently, including extremely low lake levels and poor water quality, which indicates its vulnerability against changing climatic conditions. At the same time several water usage conflicts appeared in the catchment area. Until recently, the groundwater component in the lake's water budget and the hydrogeological processes in the catchment area have not been taken into consideration. Recent hydrogeological studies, however, show groundwater discharge into the lake. Thus, further investigating this question is of high importance, hence groundwater could reduce climatic vulnerability. Our ongoing work aims at developing a model-based evaluation technique, utilizing all map-based geophysical information and time series of different satellite data products, having sufficient spatial resolution and providing information about parameters strongly connected to subsurface processes, showing up on the surface. The basic DEM raster layer is imported from Copernicus GLO-30 dataset, having vertical precision <4 m. The Region Of Interest is a rectangular part of the catchment area: 47.1–47.4N, 18.4–18.8E. The first segmentation of the ROI is done using elevation data combined with lithographic and soil type information, resulting in almost uniform Voronoi-like polygon tessellation, with cells classified by geostructure. Further refinement by land cover type is done using Sentinel-1 SAR data. Other fixed data of point and polygon layers are important terrain features, points of surface inflows, (known) water takeouts and monitoring wells. The machine learning regression model has time series of measured data at all its layers, daily input from Agárd meteorological station, like precipitation, average temperature, wind speed and relative humidity. Another important input data comes from Sentinel-2 (GREEN-NIR)/(GREEN+NIR)=NDWI spectral index, available in about weekly time steps, varying between 2 days-2 weeks. A crucial feature of all remote sensing data used here is the spatial resolution being better (10 m) or similar to the resolution of the basic DEM model. During training a graph neural network is generated dynamically from the Voronoi tessellation, where cells are nodes and physical processes between neighbouring cells give edge attributes for the graph. We use rectilinear approximations for water runoff/subsurface water exchange between cells, vertical infiltration/discharge under cells and estimated evapotranspiration from them. Learnable parameters governing the intensity of these flows are connected to geostructure and land cover classes. Parameters are optimized with time interval cross validation, with one part of the time series data being left out from optimization in each epoch and used for evaluation against target water level data. Automatic detection of spatio-temporal patterns, connected to near-surface hydrogeological processes helped visualizing and quantifying estimated physical flows. Comparison with field measurements confirmed theoretic results from MODFLOW basin modelling, proving topography as a driving factor for subsurface flows. Our model is also suitable to handle isotope tracers, and extension to deep learning model promises predictive functionality for water table level. The research is part of a project which was funded by the National Multidisciplinary Laboratory for Climate Change, (Hungary) RRF-2.3.1-21-2022-00014. LA - English DB - MTMT ER - TY - CONF AU - Kahuthu, Dennis Wambugu AU - Amimo, Meshack O. AU - Oiro, S. AU - Székely, Balázs TI - Analysis of Hydrogeological Parameters of the Nairobi Aquifer Suite Using GIS-Based Spatial Interpolation Methods T2 - EGU General Assembly 2024 : abstracts PB - European Geosciences Union (EGU) C1 - Wien PY - 2024 DO - 10.5194/egusphere-egu24-899 UR - https://m2.mtmt.hu/api/publication/34802125 ID - 34802125 AB - Groundwater resources in the Nairobi Aquifer Suite (NAS), Kenya, face significant problems largely due to rapid urbanization and the rising water demand. The depletion of groundwater resources at the local level could potentially extend to regional extents, and hence affect natural water flows. This therefore calls for the prediction of aquifer hydrogeological parameters for sustainable groundwater management. This study aims to utilize GIS-based spatial interpolation methods for the in-depth analysis of NAS hydrogeological parameters. Classical geostatistical tools are employed to develop models that can be used to accurately predict hydrogeological parameters of the NAS. Field-measurable predictors, that is, geographic position, elevation, depths and first water struck level, are used to demonstrate the efficacy of the predictive models. Data from hydrogeological measurements, geological surveys and satellite imagery are integrated during the development of the predictive models for key hydrogeological parameters, including, groundwater level, discharge, drawdown, electrical conductivity, and transmissivity. Classical geostatistical tools such as kriging and natural neighbour interpolation are applied to develop spatially explicit maps of the NAS hydrogeological parameters. The distribution of borehole data is analyzed using geostatistical tools such as trend analysis and semi variogram. Cross-validation has been performed to identify the most suitable spatial interpolation model. While, in general, the prediction worked well based on model evaluation metrics such as mean absolute error (MAE), mean squared error (MSE), root mean square error (RMSE) and coefficient of determination (R2), during the testing we observed characteristic deviations from the measured values at some locations. These differences could be due to the geological setting; however, a few outliers may appear due to yet unknown reasons. Further studies utilizing machine learning techniques are expected to develop accurate predictive models that can help in sustainable groundwater management in the NAS. The generated spatial maps provided insightful information on the spatial distribution of hydrogeological parameters in the NAS, facilitating the accurate identification of prospective locations for ideal groundwater extraction. LA - English DB - MTMT ER - TY - CONF AU - Musyimi, Peter Kinyae AU - Székely, Balázs AU - Hellen, W. Kamiri AU - Tom, Ouna AU - Weidinger, Tamás TI - Meteorological and Soil Moisture Measurements in Mount Kenya Region at Various Scales T2 - EGU General Assembly 2024 : abstracts PB - European Geosciences Union (EGU) C1 - Wien PY - 2024 DO - 10.5194/egusphere-egu24-579 UR - https://m2.mtmt.hu/api/publication/34762086 ID - 34762086 LA - English DB - MTMT ER - TY - JOUR AU - Telbisz, Tamás AU - Mari, László AU - Székely, Balázs TI - LiDAR-Based Morphometry of Dolines in Aggtelek Karst (Hungary) and Slovak Karst (Slovakia) JF - REMOTE SENSING J2 - REMOTE SENS-BASEL VL - 16 PY - 2024 IS - 5 SP - 737 SN - 2072-4292 DO - 10.3390/rs16050737 UR - https://m2.mtmt.hu/api/publication/34667579 ID - 34667579 AB - LiDAR-based digital terrain models (DTMs) represent an advance in the investigation of small-scale geomorphological features, including dolines of karst terrains. Important issues in doline morphometry are (i) which statistical distributions best model the size distribution of doline morphometric parameters and (ii) how to characterize the volume of dolines based on high-resolution DTMs. For backward compatibility, how previous datasets obtained predominantly from topographic maps relate to doline data derived from LiDAR is also examined. Our study area includes the karst plateaus of Aggtelek Karst and Slovak Karst national parks, whose caves are part of the UNESCO World Heritage. To characterize the study area, the relationships between doline parameters and topography were studied, as well as their geological characteristics. Our analysis revealed that the LiDAR-based doline density is 25% higher than the value calculated from topographic maps. Furthermore, LiDAR-based doline delineations are slightly larger and less rounded than in the case of topographic maps. The plateaus of the study area are characterized by low (5–10 km−2), moderate (10–30 km−2), and medium (30–35 km−2) doline densities. In terms of topography, the slope trend is decisive since the doline density is negligible in areas where the general slope is steeper than 12°. As for the lithology, 75% of the dolines can be linked to Wetterstein Limestone. The statistical distribution of the doline area can be well modeled by the lognormal distribution. To describe the DTM-based volume of dolines, a new parameter (k) is introduced to characterize their 3D shape: it is equal to the product of the area and the depth divided by the volume. This parameter indicates whether the idealized shape of the doline is closer to a cylinder, a bowl (calotte), a cone, or a funnel shape. The results show that most sinkholes in the study area have a transitional shape between a bowl (calotte) and a cone. LA - English DB - MTMT ER - TY - JOUR AU - Timár, Gábor AU - Jakab, Gusztáv AU - Székely, Balázs TI - A Step from Vulnerability to Resilience: Restoring the Landscape Water-Storage Capacity of the Great Hungarian Plain—An Assessment and a Proposal JF - LAND (BASEL) J2 - LAND-BASEL VL - 13 PY - 2024 IS - 2 SN - 2073-445X DO - 10.3390/land13020146 UR - https://m2.mtmt.hu/api/publication/34536076 ID - 34536076 N1 - Department of Geophysics and Space Science, Institute of Geography and Earth Science, ELTE Eötvös Loránd University, Budapest, H-1117, Hungary Department of Environmental and Landscape Geography, Institute of Geography and Earth Sciences, ELTE Eötvös Loránd University, Budapest, H-1117, Hungary Export Date: 28 March 2024 Correspondence Address: Timár, G.; Department of Geophysics and Space Science, Hungary; email: timar.gabor@ttk.elte.hu AB - The extreme drought in Europe in 2022 also hit hard the Great Hungarian Plain. In this short overview article, we summarize the natural environmental conditions of the region and the impact of river control works on the water-retention capacity of the landscape. In this respect, we also review the impact of intensive agricultural cultivation on soil structure and on soil moisture in light of the meteorological elements of the 2022 drought. The most important change is that the soil stores much less moisture than in the natural state; therefore, under the meteorological conditions of summer 2022, the evapotranspiration capacity was reduced. As a result, the low humidity in the air layers above the ground is not sufficient to trigger summer showers and thunderstorms associated with weather fronts and local heat convection anymore. Our proposed solution is to restore about one-fifth of the area to the original land types and usage before large-field agriculture. Low-lying areas should be transformed into a mosaic-like landscape with good water supply and evapotranspiration capacity to humidify the lower air layers. Furthermore, the unfavorable soil structure that has resulted from intensive agriculture should also be converted into more permeable soil to enhance infiltration. LA - English DB - MTMT ER - TY - CHAP AU - Musyimi, Peter Kinyae AU - Székely, Balázs AU - Weidinger, Tamás ED - Kis, Anna ED - Pongrácz, Rita TI - Meteorological measurements in Mount Kenya region, importance of quality control, preparatory steps and calibration T2 - Aktuális doktori kutatások a levegőkémia, a klímaváltozás és a meteorológia témakörében PB - ELTE TTK Meteorológiai Tanszék CY - Budapest SN - 9789634896333 T3 - Egyetemi meteorológiai füzetek, ISSN 0865-7920 ; 35. PY - 2023 SP - 155 EP - 169 PG - 15 DO - 10.31852/EMF.35.2023.155.169 UR - https://m2.mtmt.hu/api/publication/34325400 ID - 34325400 LA - English DB - MTMT ER - TY - JOUR AU - Sahbeni, Ghada AU - Székely, Balázs AU - Musyimi, Peter Kinyae AU - Timár, Gábor AU - Sahajpal, Ritvik TI - Crop Yield Estimation Using Sentinel-3 SLSTR, Soil Data, and Topographic Features Combined with Machine Learning Modeling: A Case Study of Nepal JF - AgriEngineering J2 - AgriEngineering VL - 5 PY - 2023 IS - 4 SP - 1766 EP - 1788 PG - 23 SN - 2624-7402 DO - 10.3390/agriengineering5040109 UR - https://m2.mtmt.hu/api/publication/34190861 ID - 34190861 N1 - Department of Geophysics and Space Science, Institute of Geography and Earth Sciences, ELTE Eötvös Loránd University, Pázmány Péter stny. 1/C, Budapest, H-1117, Hungary Department of Meteorology, Institute of Geography and Earth Sciences, ELTE Eötvös Loránd University, Pázmány Péter stny. 1/A, Budapest, H-1117, Hungary Department of Humanities and Languages, Karatina University, P.O. Box 1957-10101, Karatina, Kenya NASA-Harvest, Department of Geographical Sciences, University of Maryland, College Park, MD 20742, United States Export Date: 4 January 2024 Correspondence Address: Sahbeni, G.; Department of Geophysics and Space Science, Pázmány Péter stny. 1/C, Hungary; email: gsahbeni@caesar.elte.hu AB - Effective crop monitoring and accurate yield estimation are fundamental for informed decision-making in agricultural management. In this context, the present research focuses on estimating wheat yield in Nepal at the district level by combining Sentinel-3 SLSTR imagery with soil data and topographic features. Due to Nepal’s high-relief terrain, its districts exhibit diverse geographic and soil properties, leading to a wide range of yields, which poses challenges for modeling efforts. In light of this, we evaluated the performance of two machine learning algorithms, namely, the gradient boosting machine (GBM) and the extreme gradient boosting (XGBoost). The results demonstrated the superiority of the XGBoost-based model, achieving a determination coefficient (R2) of 0.89 and an RMSE of 0.3 t/ha for training, with an R2 of 0.61 and an RMSE of 0.42 t/ha for testing. The calibrated model improved the overall accuracy of yield estimates by up to 10% compared to GBM. Notably, total nitrogen content, slope, total column water vapor (TCWV), organic matter, and fractional vegetation cover (FVC) significantly influenced the predicted values. This study highlights the effectiveness of combining multi-source data and Sentinel-3 SLSTR, particularly proposing XGBoost as an alternative tool for accurately estimating yield at lower costs. Consequently, the findings suggest comprehensive and robust estimation models for spatially explicit yield forecasting and near-future yield projection using satellite data acquired two months before harvest. Future work can focus on assessing the suitability of agronomic practices in the region, thereby contributing to the early detection of yield anomalies and ensuring food security at the national level. LA - English DB - MTMT ER - TY - CONF AU - Musyimi, Peter Kinyae AU - Mendyl, Abderrahmane AU - Agustiyara, Agustiyara AU - Székely, Balázs AU - Weidinger, Tamás TI - Hourly reference evapotranspiration analysis using synoptic meteorological measurements and ERA5 reanalysis data from Kenyan Counties T2 - EMS Annual Meeting Abstracts PY - 2023 DO - 10.5194/ems2023-331 UR - https://m2.mtmt.hu/api/publication/34128935 ID - 34128935 LA - English DB - MTMT ER - TY - CHAP AU - Agustiyara, Agustiyara AU - Székely, Balázs AU - Nurmandi, Achmad AU - Musyimi, Peter Kinyae ED - Salvendy, Gavriel ED - Ntoa, Stavroula ED - Antona, Margherita ED - Stephanidis, Constantine TI - Remote Sensing Applied for Land Use Change Assessment and Governance in Riau-Indonesia T2 - HCI International 2023 Posters PB - Springer Nature Switzerland AG CY - Cham SN - 9783031360015 T3 - Communications in Computer and Information Science, ISSN 1865-0929 ; 1835. PY - 2023 SP - 441 EP - 448 PG - 8 DO - 10.1007/978-3-031-36001-5_56 UR - https://m2.mtmt.hu/api/publication/34058765 ID - 34058765 N1 - Export Date: 1 December 2023 Correspondence Address: Agustiyara, A.; Department of Geophysics and Space Science, Hungary; email: agustiyara@student.elte.hu AB - Remote sensing offers the potential to provide up-to-date information on changes in forestry areas over large areas. Its application makes it possible to make assessments related to land use change. This research aims to assess whether land change using remote sensing can provide an efficient alternative, both in terms of cost and time, including improving forest governance policy support. Remote sensing and forest governance are state-of-the-art in this research for the development of knowledge from in-depth data analysis. This study was conducted in Bengkalis-Riau Province, Indonesia because, the regency has become the most vulnerable region for forest fires since 2013 and the province has experienced growing pressure from an expanding palm oil industry. It has the largest tropical peatland area and palm oil plantation in Indonesia. The use of remote sensing data methods improved the sensitivity of detecting classified forest cover, providing a better understanding of changes that are usually difficult to map, including fires, smallholders and industrial scale of agricultural areas, peatland cover, wetlands, and barren forest land. Both smallholder and industrial agricultural areas are also better detected. The result from Sentinel data indicate forest, and land cover changes after evaluation, which focuses on the spatial, spectral, and temporal resolution of the imagery. The cover of land use change generated by remote sensing data shows the classification of land conditions in the study area, ranging from cultivated land, bare soil, forestry, oil palm plantations, and peatlands within the plantation area. Integration of artificial intelligence will be further explored. LA - English DB - MTMT ER - TY - CHAP AU - Musyimi, Peter Kinyae AU - Helga, Chauke AU - Székely, Balázs AU - Weidinger, Tamás ED - Hrvojević, MPD TI - Decadal Geographical Variability of Temperature and Precipitation-based Climate Extremes in Different Climate Regions of Kenya T2 - International Conference on Hydro-Climate Extremes and Society PB - University of Novi Sad, Faculty of Sciences CY - Novi Sad SN - 9788670316225 PY - 2023 SP - 45 EP - 46 PG - 2 UR - https://m2.mtmt.hu/api/publication/34043412 ID - 34043412 LA - English DB - MTMT ER -