Carmen Maftei

This study analyzes the performance of several forecasting models applied to Standardized Precipitation Index (SPI) time series for the 1967–2021 period, using data from six weather stations in southeastern Romania. Four temporal aggregation scales (SPI3, SPI6, SPI12, and SPI24) were considered for one-step-ahead (one-month) forecasting. The evaluated models included Random Forest, XGBoost, Support Vector Regression (SVR), N-BEATS (generic variant), and SARIMA (Seasonal Autoregressive Integrated Moving Average), covering machine learning, deep learning, and statistical approaches. To ensure a rigorous out-of-sample evaluation, SPI distribution parameters were estimated solely based on the pre-test calibration period and held constant when calculating SPI values for the independent test period, thereby preventing information leakage from the preprocessing stage. Forecast performance was assessed using RMSE, MAE, NSE, and a skill score relative to a persistence benchmark, while differences in forecast accuracy were statistically evaluated using the Diebold–Mariano test. The results revealed a clear scale-dependent pattern: absolute forecast accuracy increased with the SPI aggregation scale—alongside more pronounced temporal persistence—though this did not necessarily translate into superior skill compared to the persistence model. Among the data-driven models, SVR demonstrated the most robust performance, particularly at shorter timescales, while N-BEATS also provided competitive forecasts without consistently outperforming SVR. SARIMA showed the clearest gains over the persistence model at the SPI12 and SPI24 scales. The maximum NSE increased from 0.61 for SPI3 and 0.81 for SPI6, both obtained with SVR, to 0.94 for SPI12 (obtained with SARIMA) and 0.97 for SPI24 (obtained with SARIMA, N-BEATS and SVR). In conclusion, the results demonstrate that the SPI accumulation scale, temporal structure, and performance relative to persistence must be considered simultaneously when selecting forecasting models for drought prediction.

The aim of this study is to develop an integrated methodological framework for assessing flood susceptibility, agricultural land vulnerability, and landscape structure using Geographic Information Systems (GIS) and the Analytic Hierarchy Process (AHP), thereby contributing to the integrated assessment of these components at the scale of an experimental micro-watershed. The research was based on a 2.5 m resolution Digital Elevation Model (DEM) and detailed data obtained through field investigations. Flood susceptibility was modeled by integrating geomorphological, hydrological, pedological, and land-use factors. The results highlight the predominance of the moderate susceptibility class (64.53%), while the high and very high susceptibility classes account for 14.22% of the watershed area and are concentrated in the vicinity of the drainage network and in low-lying sectors. The assessment of land vulnerability revealed differences among land-use categories. Hayfields encompass the largest area exposed to high and very high flood susceptibility classes, whereas forested areas are predominantly associated with low and moderate susceptibility classes. The landscape structure analysis revealed very strong correlations between the exposed area and the landscape metrics used to characterize landscape structure. The proposed integrated methodological framework provides a transferable approach for the simultaneous assessment of flood susceptibility, agricultural land vulnerability, and landscape structure in micro-watersheds with similar characteristics.

Landfill leachate represents a long-term source of contamination that may significantly affect groundwater and receiving water bodies through the migration of organic, inorganic, and toxic pollutants. This study evaluated the long-term migration of landfill leachate and its potential environmental impacts using the LandSim Release 2 probabilistic software model applied to two municipal waste landfills in the Republic of Serbia: the regional sanitary landfill “Gigoš” in Jagodina and the sanitary landfill “Meteris” in Vranje. The modelling framework integrated laboratory leachate analyses, hydrogeological conditions, engineered barrier system characteristics, and receptor-oriented contaminant transport assessment. Model validation was performed through comparison of simulated and laboratory-measured concentrations. Two scenarios were analyzed for each site: an engineered sanitary landfill scenario with a functional containment system and a conservative barrier-failure scenario representing complete loss of engineered barrier functionality. Ten representative leachate parameters were included, covering nitrogen compounds, inorganic ions, toxic substances, and heavy metals/metalloids. The results showed that engineered protection systems significantly delay contaminant migration and reduce receptor concentrations, while barrier-failure conditions lead to earlier pollutant breakthrough and higher environmental risk. The simulations demonstrated that under the engineered sanitary landfill scenario, receptor concentrations of all analyzed contaminants remained below the corresponding maximum allowable concentrations, with contaminant migration occurring only after several centuries. In contrast, the conservative barrier-failure scenario resulted in substantially earlier contaminant breakthrough, with nitrogen compounds and phenols representing the greatest environmental concern due to their rapid migration and exceedance of regulatory thresholds, while the “Meteris” landfill generally exhibited higher receptor concentrations than the “Gigoš” landfill. These findings highlight the importance of predictive modelling and long-term monitoring for sustainable landfill management and groundwater protection.

Global anthropogenic activities and conjoint climate extremes are fundamentally disrupting closely coupled hydroecological systems, altering catchment-scale water balances, biogeochemical fluxes, and the resilience of ecosystem services

Flooding is a recurrent natural hazard in Romania, causing significant socio-economic impacts. Historical data highlights the severity of floods, particularly the unprecedented flood of 1926. Between 1960 and 2010, Romania experienced over 400 major floods, which significantly impacted its infrastructure and population. Particularly, the floods in 2005 and 2006 affected over 1.5 million people, resulting in 93 deaths and causing damages exceeding EUR 2 billion. In compliance with the Floods Directive, EU member states must assess and map flood hazards and risks. This study aims to develop a frequency analysis to determine discharges as a predictive indicator for different hazard levels: frequent events (10-year return period), medium probability events (100-year return period), and extreme events. The Casimcea catchment in central Dobrogea, drained by the Casimcea River into Lake Tasaul, serves as the study area. The annual maximum discharge data analysis, conducted through frequency analysis and the ELECTRE method, indicates that EV3-Min-Weibull, L-moments, and GEV-Min (L-moments) are the most effective probability density functions (PDFs). To conclude, although a single PDF model cannot be determined for the Casimcea River and its tributaries, it contributes to predictive modeling efforts.

This paper analyzes the chemical evolution of Nuntași-Tuzla Lake (Romania) in the context of human intervention. Situated on the shore of the Black Sea, approximately 35 km north of Constanța, Nuntași-Tuzla Lake is part of the Razim–Sinoe Lake complex and a component of the Danube Delta Biosphere Reserve. This area has undergone significant transformations over the past 120 years: canalization of the connecting channels with the St. George arm, construction of polders for agriculture, closure of the connections to the Black Sea, and construction of the Razim–Sinoe irrigation system. After the irrigation system stopped working (around 2000), due to the isolation of the lake and the low flow coming from the two rivers that supply the lake with fresh water, it completely dried up in 2020. All these interventions have led to the ecological, hydrological, and chemical deterioration of the lake’s water. The main effects are (i) a decrease in water salinity and (ii) reduction in the production of sapropelic mud as the salinity decreases due to the influx of fresh water.

Drought raises significant challenges and consequences in the socioeconomic environment in Dobrogea, Romania. This research aimed to assess the spatiotemporal dynamics of agrometeorological droughts from 2001 to 2021 using a multi-index approach that includes the Vegetation Health Index (VHI) and Standardized Precipitation Evapotranspiration Index (SPEI). Severe-to-extreme drought events were detected in 2001, 2007, 2012, 2015, 2016, 2019, and 2020, when temperatures in the area reached as high as 40.91 °C. Regarding area coverage, 2012 and 2020 were the worst drought years, with 66% and 71% of the region affected. Mild and moderate droughts were consistently identified across almost the entire period, while normal wet conditions were indicated in 2004–2006. The spatial analysis and the drought frequency maps revealed that the central, southern, and northwestern areas were particularly vulnerable, underlining the need for targeted drought mitigation measures. The trend analysis results indicated a nonuniform spatial feature of the negative (drying)/positive (wetting) trends at the regional level, with statistically significant trends identified only over small areas. Further results showed a robust relationship among the VHI and SPEI, particularly on 1-month and seasonal timescales. The extended correlation analysis results showed very strong positive relationships among all the vegetation indices, positive relations with rainfall, and strong negative ties with land surface temperature. Moreover, the seasonal VHI proved to be effective for drought monitoring across areas with diverse crop types. The results we obtained are consistent with previous studies on the incidence of drought in the area and hold practical significance for decision-makers responsible for drought management planning within Dobrogea, including setting up an early warning system using the VHI.

This paper discusses a study that examined the severity of droughts and their changes in the Dobrogea region in southeastern Romania between 2001 and 2021 and develops a high-resolution (1 km) Composite Drought Index (CDI) dataset. To explore the effectiveness of the index, we carried out a correlation analysis between the CDI, the Standardized Precipitation Index (SPI), and the Standardized Precipitation Evapotranspiration Index (SPEI), which shows a strong positive relationship among these indices. Analysis of the CDI time series reveals an increase in drought frequency for the study period, due to high temperature and below-normal rainfall. Most parts of the region were affected by moderate, severe, or extreme droughts, except for the years 2002–2005 and 2013. The worst drought events were in 2011, 2012, and 2020, when the region was under severe land surface temperature stress, with values up to 39.13 °C. The central and northern areas of the region had the longest period of drought, at 22 months, which started in 2018 and culminated in 2020 when extreme drought covered over 70% of the region. Another major event was in 2015 when 95% of the region experienced severe drought. These results show the potential of the CDI as one of the significant indices in the assessment of drought and provide useful insights into drought monitoring in the future. More than that, we consider that the GPM IMERG satellite product can be used in the implementation of Drought Management Plans in Dobrogea in order to calculate drought indices derived from remote sensing data.

Drought poses a significant risk in many parts of the world, especially in regions reliant on agriculture. Evaluating this risk is an essential step in preventing and reducing its impact. In this context, we assess the drought intensity at six sites in Constanța County (Romania) using the de Martonne aridity index. The risk of aridity and vulnerability to drought were evaluated by the Drought Hazard Index (DHI) and Drought Risk Index (DRI), computed based on the Standardized Precipitation Index (SPI). The de Martonne index indicates a variation between the slightly arid and semi-arid climates for Adamclisi station, with periodic changes from semi-arid to arid. At Cernavodă station, we notice a passage from an arid period towards a moderately humid one (in 2005), followed by a movement in the opposite direction to the limit of the arid zone (in 2011), and a return inside the “limits” of the semi-arid to moderately arid climate. A similar variation for 2000–2018 is noticed at Medgidia, Hârșova, and Mangalia. DRI classifies two stations in the low risk to drought category and one in the moderate risk to drought class. The other two locations experience a high or very high risk of drought. The drought intensities varied in the intervals 0.503–1.109 at Constanța, 0.473–1.363 at Mangalia, 0.511–1.493 at Adamclisi, 0.438–1.602 at Hârșova, 0.307–1.687 at Medgidia, and 0.463–1.307 at Cernavodă, and the prolonged drought periods were over 99 months at all stations.

Air pollution poses one of the greatest dangers to public well-being. This article outlines a study conducted in the Central Romania Region regarding the health risks associated with particulate matter (PM) of two sizes, viz., PM₁₀ and PM_2.5. The methodology used consists of the following: (i) an analysis of the effects of PM pollutants, (ii) an analysis of total mortality and cardiovascular-related mortality, and (iii) a general health risk assessment. The Central Region of Romania is situated in the Carpathian Mountains’ inner arch (consisting of six counties). The total population of the region under investigation is about 2.6 million inhabitants. Health risk assessment is calculated based on the relative risk (RR) formula. During the study period, our simulations show that reducing these pollutants’ concentrations below the new WHO guidelines (2021) will prevent over 172 total fatalities in Brasov alone, as an example. Furthermore, the potential benefit of reducing annual PM_2.5 levels on total cardiovascular mortality is around 188 persons in Brasov. Although health benefits may also depend upon other physiological parameters, all general health indicators point towards a significant improvement in overall health by a general reduction in particulate matter, as is shown by the toxicity assessment of the particulate matter in the region of interest. The modality can be applied to other locations for similar studies.