Search

Crop condition mapping and yield loss detection are highly relevant scientific fields due to their economic importance. Here, we report a new, robust six-category crop condition mapping methodology based on five vegetation indices (VIs) using Sentinel-2 imagery at a 10 m spatial resolution. We focused on maize, the most drought-affected crop in the Carpathian Basin, using three selected years of data (2017, 2022, and 2023). Our methodology was validated at two different spatial scales against independent reference data. At the parcel level, we used harvester-derived precision yield data from six maize parcels. The agreement between the yield category maps and those predicted from the crop condition time series by our Random Forest model was 84.56%, while the F1 score was 0.74 with a two-category yield map. Using a six-category yield map, the accuracy decreased to 48.57%, while the F1 score was 0.42. The parcel-level analysis corroborates the applicability of the method on large scales. Country-level validation was conducted for the six-category crop condition map against official county-scale census data. The proportion of areas with the best and worst crop condition categories in July explained 64% and 77% of the crop yield variability at the county level, respectively. We found that the inclusion of the year 2022 (associated with a severe drought event) was important, as it represented a strong baseline for the scaling. The study’s novelty is also supported by the inclusion of damage claims from the Hungarian Agricultural Risk Management System (ARMS). The crop condition map was compared with these claims, with further quantitative analysis confirming the method’s applicability. This method offers a cost-effective solution for assessing damage claims and can provide early yield loss estimates using only remote sensing data.

Dissolved organic matter (DOM) is a key soil quality property, indicative of the organic matter stored in the soil, which may also be a function of temporal variation. This study examines whether DOM is a robust property of the soil, controlling fertility, or if it may change with time. Altogether eight sets of soil samples were collected in 2018 and 2019 from the cultivated topsoil (0–10 cm) of cropland and from a nearby grassland near Martonvásár, Hungary. The study sites were characterized by Chernozem soil and were part of a long-term experimental project comparing the effects of manure application and fertilization to the control under maize and wheat monocultures. DOM was extracted from the samples with distilled water. The dissolved organic carbon (DOC), total dissolved nitrogen (DN), biological index (BIX), fluorescence index (FI), humification index (HIX), carbon nitrogen (C/N) ratio and specific ultraviolet absorbance at 254 nm (SUVA254) index were studied in the arable soils, and the results showed that all the DOM samples were humified, suggesting relevant microbiological contributions to the decomposition of OM and its conversion into more complex molecules (FI = 1.2–1.5, BIX = ~0.5, and HIX = ~0.9). Temporal variations were detected only for the permanent grassland where higher DOM concentration was found in spring. This increased DOM content mainly originated from humified, solid phase associated, recalcitrant OM. In contrast, there were no differences among fertilization treatments and sampling dates under cropfield conditions. Moreover, climatic conditions were not proven as a general ruler of DOM properties. Therefore, momentary DOM alone is not necessarily the direct property of soil organic matter under cropfield conditions. The application of this measure needs further details of sampling conditions to achieve adequate comparability.

Previous studies have suggested that a major part of the observed variability in vegetation state might be associated with variability in climatic drivers during relatively short periods within the year. Identification of such critical climate periods, when a particular climate variable most likely has a pronounced influence on the vegetation state of a particular ecosystem, becomes increasingly important in the light of climate change. In this study, we present a method to identify critical climate periods for eight different semi-natural ecosystem categories in Hungary, in Central Europe. The analysis was based on the moving-window correlation between MODIS NDVI/LAI and six climate variables with different time lags during the period 2000–2020. Distinct differences between the important climate variables, critical period lengths, and direction (positive or negative correlations) have been found for different ecosystem categories. Multiple linear models for NDVI and LAI were constructed to quantify the multivariate influence of the environmental conditions on the vegetation state during the late summer. For grasslands, the best models for NDVI explained 65–87% variance, while for broad-leaved forests, the highest explained variance for LAI was up to 50%. The proposed method can be easily implemented in other geographical locations and can provide essential insight into the functioning of different ecosystem types.

Soil organic carbon (SOC) is a mandatory pool in national inventory reports on greenhouse gas (GHG) emissions and removals to the UNFCCC. Hence, its accurate assessment is important. Modelling SOC changes for national GHG reports is encouraged, but the uncertainty related to this pool still presents a significant challenge; thus, verifying modelling results with field observations is essential. We used the process-based model Biome-BGCMuSo and assessed its suitability for use in Croatia’s GHG reporting. We modelled SOC stocks in the top 30 cm of the mineral soil layer (SOC30) for four different land-use (LU) categories (Deciduous/Coniferous Forest, Grassland and Annual Cropland) distributed in three biogeographical regions (Alpine, Continental and Mediterranean) and compared them with results of a national soil survey. A total of 573 plot level simulations were undertaken and results were evaluated at three stratification levels (LU, LU × biogeographical region, and plot). The model reproduced the overall country mean of SOC30 with no overall bias, and showed good performance at the LU level with no significant (p < 0.05) difference for all LUs except Deciduous Forest (11% overestimation). At finer stratifications, the model performance considerably worsened. Further model calibration, improvement and testing, as well as repeated soil survey are needed in order to assess the changes in SOC30 and to evaluate the potential of the Biome-BGCMuSo model for use in GHG reporting.

The evaluation of general circulation models (GCM) is a fundamental step in climate research in terms of both quality assurance/quality control and realistic representation of the dynamics of the atmospheric flows in the future projections. In this paper, a statistical method is introduced to evaluate GCMs with respect to teleconnection patterns in the winter 500 hPa geopotential height field over the Northern Hemisphere (NH). The procedure uses the combination of negative extrema method and receiver operating characteristic (ROC) curve analysis. The proposed method is demonstrated using selected general circulation models (GCMs) disseminated by the CMIP5 project. The ERA-20C reanalysis was used as a reference, supported by the NCEP/NCAR R1 reanalysis. The proposed method enables us to track changes in the geographical positions of the action centers (ACs); therefore, to detect improvement/deterioration in the GCM performance with time. It was found that the majority of the GCMs reproduce prominent teleconnections of the NH but fail to capture the eastward shift of the ACs over the Pacific Ocean in the last decades of the 20th century. The GCMs reproduce teleconnections with stronger correlations over the north-western part of the Atlantic Ocean compared to the reanalyses. The construction of mobile teleconnection indices is proposed to gain further insight into the performance of the models and to support a regional-scale analysis. The method can be easily applied to the recent CMIP6 simulations.

Accurate estimation of the timing of intensive spring leaf growth initiation at mid and high latitudes is crucial for improving the predictive capacity of biogeochemical and Earth system models. In this study, we focus on the modeling of climatological onset of spring leaf growth in Central Europe and use three spring phenology models driven by three meteorological datasets. The MODIS-adjusted NDVI3g dataset was used as a reference for the period between 1982 and 2010, enabling us to study the long-term mean leaf onset timing and its interannual variability (IAV). The performance of all phenology model–meteorology database combinations was evaluated with one another, and against the reference dataset. We found that none of the constructed model–database combinations could reproduce the observed start of season (SOS) climatology within the study region. The models typically overestimated IAV of the leaf onset, where spatial median SOS dates were best simulated by the models based on heat accumulation. When aggregated for the whole study area, the complex, bioclimatic index-based model driven by the CarpatClim database could capture the observed overall SOS trend. Our results indicate that the simulated timing of leaf onset primarily depends on the choice of model structure, with a secondary contribution from the choice of the driving meteorological dataset.

Atmospheric teleconnections are characteristic to the climate system and exert major impacts on the global and regional climate. Accurate representation of teleconnections by general circulation models (GCMs) is indispensable given their fundamental role in the large scale circulation patterns. In this study a statistical method is introduced to evaluate historical GCM outputs of the fifth phase of the Coupled Model Intercomparison Project (CMIP5) with respect to teleconnection patterns. The introduced method is based on the calculation of correlations between gridded time series of the 500 hPa geopotential height fields in the Northern Hemisphere. GCMs are quantified by a simple diversity index. Additionally, potential action centers of the teleconnection patterns are identified on which the local polynomial regression model is fitted. Diversity fields and regression curves obtained from the GCMs are compared against the NCEP/NCAR Reanalysis 1 and the ERA-20C reanalysis datasets. The introduced method is objective, reproducible, and reduces the number of arbitrary decisions during the analysis. We conclude that major teleconnection patterns are positioned in the GCMs and in the reanalysis datasets similarly, however, spatial differences in their intensities can be severe in some cases that could hamper the applicability of the GCM results for some regions. Based on the evaluation method, best-performing GCMs can be clearly distinguished. Evaluation of the GCMs based on the introduced method might help the modeling community to choose GCMs that are the most applicable for impact studies and for regional downscaling exercises.

Multi-species grasslands are reservoirs of biodiversity and provide multiple ecosystem services, including fodder production and carbon sequestration. The provision of these services depends on the control exerted on the biogeochemistry and plant diversity of the system by the interplay of biotic and abiotic factors, e.g., grazing or mowing intensity. Biogeochemical models incorporate a mechanistic view of the functioning of grasslands and provide a sound basis for studying the underlying processes. However, in these models, the simulation of biogeochemical cycles is generally not coupled to simulation of plant species dynamics, which leads to considerable uncertainty about the quality of predictions. Ecological models, on the other hand, do account for biodiversity with approaches adopted from plant demography, but without linking the dynamics of plant species to the biogeochemical processes occurring at the community level, and this hampers the models’ capacity to assess resilience against abiotic stresses such as drought and nutrient limitation. While setting out the state-of-the-art developments of biogeochemical and ecological modelling, we explore and highlight the role of plant diversity in the regulation of the ecosystem processes underlying the ecosystems services provided by multi-species grasslands. An extensive literature and model survey was carried out with an emphasis on technically advanced models reconciling biogeochemistry and biodiversity, which are readily applicable to managed grasslands in temperate latitudes. We propose a roadmap of promising developments in modelling.

Remote sensing provides invaluable insight into the dynamics of vegetation with global coverage and reasonable temporal resolution. Normalized Difference Vegetation Index (NDVI) is widely used to study vegetation greenness, production, phenology and the responses of ecosystems to climate fluctuations. The extended global NDVI3g dataset created by Global Inventory Modeling and Mapping Studies (GIMMS) has an exceptional 32 years temporal coverage. Due to the methodology that was used to create NDVI3g inherent noise and uncertainty is present in the dataset. To evaluate the accuracy and uncertainty of application of NDVI3g at regional scale we used Collection-6 data from the MODerate resolution Imaging Spectroradiometer (MODIS) sensor on board satellite Terra as a reference. After noise filtering, statistical harmonization of the NDVI3g dataset was performed for Central Europe based on MOD13 NDVI. Mean seasonal NDVI profiles, start, end and length of the growing season, magnitude and timing of peak NDVI were calculated from NDVI3g (original, noise filtered and harmonized) and MODIS NDVI and compared with each other. NDVI anomalies were also compared and evaluated using simple climate sensitivity metrics. The results showed that (1) the original NDVI3g has limited applicability in Central Europe, which was also implied by the significant disagreement between the NDVI3g and MODIS NDVI datasets; (2) the harmonization of NDVI3g with MODIS NDVI is promising since the newly created dataset showed improved quality for diverse vegetation metrics. For NDVI anomaly detection NDVI3g showed limited applicability, even after harmonization. Climate–NDVI relationships are not represented well by NDVI3g. The presented results can help researchers to assess the expected quality of the NDVI3g-based studies in Central Europe.