Physically based vs. data-driven models for streamflow and reservoir volume prediction at a data-scarce semi-arid basin.

Physically based or data-driven models can be used for understanding basinwide hydrological processes and creating predictions for future conditions. Physically based models use physical laws and principles to represent hydrological processes. In contrast, data-driven models focus on input-output relationships. Although both approaches have found applications in hydrology, studies that compare these approaches are still limited for data-scarce, semi-arid basins with altered hydrological regimes. This study aims to compare the performances of a physically based model (Soil and Water Assessment Tool (SWAT)) and a data-driven model (Nonlinear AutoRegressive eXogenous model (NARX)) for reservoir volume and streamflow prediction in a data-scarce semi-arid region. The study was conducted in the Tersakan Basin, a semi-arid agricultural basin in Türkiye, where the basin hydrology was significantly altered due to reservoirs (Ladik and Yedikir Reservoir) constructed for irrigation purposes. The models were calibrated and validated for streamflow and reservoir volumes. The results show that (1) NARX performed better in the prediction of water volumes of Ladik and Yedikir Reservoirs and streamflow at the basin outlet than SWAT (2). The SWAT and NARX models both provided the best performance when predicting water volumes at the Ladik reservoir. Both models provided the second best performance during the prediction of water volumes at the Yedikir reservoir. The model performances were the lowest for prediction of streamflow at the basin outlet (3). Comparison of physically based and data-driven models is challenging due to their different characteristics and input data requirements. In this study, the data-driven model provided higher performance than the physically based model. However, input data used for establishing the physically based model had several uncertainties, which may be responsible for the lower performance. Data-driven models can provide alternatives to physically-based models under data-scarce conditions.

Impact of Natural Organic Matter Competition on the Adsorptive Removal of Acetochlor and Metolachlor from Low-Specific UV Absorbance Surface Waters.

Although activated carbon adsorption is a very promising process for the removal of organic compounds from surface waters, the removal performance for nonionic pesticides could be adversely affected by co-occurring natural organic matter. Natural organic matter can compete with pesticides during the adsorption process, and the size of natural organic matter affects the removal of pesticides, as low-molecular-weight organics directly compete for adsorbent sites with pesticides. This study aims to investigate the competitive impact of low-molecular-weight organics on the adsorptive removal of acetochlor and metolachlor by four commercial powdered activated carbons. The adsorption features of selected powdered activated carbons were evaluated in surface water samples collected from the influent stream of the filtration process having 2.75 mg/L organic matter and 0.87 L/mg-m specific UV absorbance. The adsorption kinetics and capacities were examined by employing pseudo-first-order, pseudo-second-order, and intraparticle diffusion kinetic models and modified Freundlich and Langmuir isotherm models to the experimental data. The competitive removal of acetochlor and metolachlor in the presence of natural organic matter was evaluated for varied powdered activated carbon dosages on the basis of UV and specific UV absorbance values of adsorbed organic matter. The adsorption data were well represented by the modified Freundlich isotherm, as well as pseudo-second-order kinetics. The maximum organic matter adsorption capacities of the modified Freundlich isotherm were observed to be 120.6 and 127.2 mg/g by Norit SX Ultra and 99.5 and 100.6 mg/g by AC Puriss for acetochlor- and metolachlor-containing water samples, respectively. Among the four powdered activated carbons, Norit SX Ultra and AC Puriss provided the highest natural organic matter removal performances with 76 and 72% and 71 and 65% for acetochlor- and metolachlor-containing samples, respectively. Similarly, Norit SX Ultra and AC Puriss were very effective for adsorbing aromatic organics with higher than 80% specific UV absorbance removal efficiency. Metolachlor was almost completely removed by higher than 98% by Norit SX Ultra, Norit SX F Cat, and AC Puriss, even at low adsorbent dosages. However, an adsorbent dose of 100 mg/L and above should be added for all powdered activated carbons, except for Norit SX F Cat, for achieving an acetochlor removal performance of higher than 98%. The competition between low-molecular-weight organics (low-specific UV absorbance) and acetochlor and metolachlor was more apparent at low adsorbent dosages (10-75 mg/L).

Multi-temporal assessment of land surface temperatures as an indicator of land use/cover changes and climate variability in the Develi Basin, Turkey.

Land surface temperature (LST) is an important parameter that reflects land surface processes of water and energy balance and has been used in assessment of land use/cover changes. However, the use of LST in monitoring changes in non-urban areas such as agricultural areas and wetlands is still limited. In this study, we aim to determine the spatial and temporal changes in LST in a semi-arid agricultural basin in Turkey (Develi Basin), where land use/cover and climatic conditions showed considerable variability since 1980s. Irrigated agriculture have expanded in the basin since 1987, after the construction of a large irrigation project. The basin hosts an internationally important wetland, called the Sultan Marshes, affected negatively by irrigation expansion. The study covers a 39-year period from 1984 to 2022. Four Landsat Thematic Mapper (TM) images acquired in 1984, 1987, 2003, and 2007 and two Landsat 8 OLI/TIRS images acquired in 2014 and 2022 were used in the analyses. The land use/cover changes were evaluated based on Normalized Difference Vegetation Index (NDVI). LST was estimated through top-of-atmosphere brightness temperature from thermal bands of Landsat images. Climate variability from 2014 to 2022 was analyzed with statistical methods. The results indicated that Develi Basin faced both spatial and temporal land use/cover changes. The area covered with natural steppe vegetation and water bodies decreased in the basin. In contrast, the sparsely and densely vegetated soil covers, which mostly denote agricultural areas, increased. Changes in LST values were observed from 1984 to 2022 as a result of climatic factors and land use/cover changes. LST changes were variable across different land use/cover types; LST decreased in irrigated areas and increased in lakes that went dry over years. LST changes proved useful for evaluating land use/cover changes and climatic variations in agricultural basins.

Modelling climate change impacts at a drinking water reservoir in Turkey and implications for reservoir management in semi-arid regions.

Climate change can have severe impacts on the water availability in semi-arid regions. In this study, we assessed the impact of climatic changes on water availability in the Altinapa Reservoir Watershed, located in the Konya province, south-central Turkey. Altinapa Reservoir supplies drinking water to Konya, a city of about 2 million population. We investigated possible changes in streamflow and reservoir storage over 2021-2098 under two representative concentration pathway scenarios (RCP4.5 and RCP8.5) developed based on GFDL-ESM2M, HadGEM2-ES, and MPI-ESM-MR global circulation models. We used a physically based model (SWAT-Soil and Water Assessment Tool) for understanding the hydrologic response of the basin to climatic changes. Results show that upward trends in air temperatures in the range of 0.01-0.04 °C/year and 0.005-0.06 °C/year are expected from 2021 to 2098 under the RCP4.5 and RCP8.5 scenarios, respectively. According to the HadGEM2-ES model, precipitation and streamflow would show a downward trend at a rate of 0.96 mm/year and 0.007 m3/s/year under the RCP4.5 scenario and at a rate of 1.62 mm/year and 0.01 m3/s/year under the RCP8.5 scenario, respectively. GFDL-ESM2M and MPI-ESM-MR models project upward trends in precipitation and streamflow under the RCP4.5 scenario (in the range of 0.64-1.28 mm/year and 0.0003-0.006 m3/s/year, respectively), and downward trends under the RCP8.5 scenario (in the range of 0.47-0.76 mm/year and 0.0015-0.003 m3/s/year, respectively). Reservoir storage is projected to increase slightly according to GFDL-ESM2M model and decrease according to the HadGEM2-ES, and MPI-ESM-MR models under both scenarios. Precipitation, streamflow, and reservoir storage predictions of GFDL-ESM2M and MPI-ESM-MR models are considerably lower than those observed in the basin in recent decades, showing that water resources will decrease in the future. The changes in water withdrawal patterns could cause further reductions in water availability. Good resilience to climate change can be achieved by a flexible water management system and by reducing water consumption and water losses in the watershed and from the reservoirs.

Spatial/temporal distribution and multi-pathway cancer risk assessment of trihalomethanes in low TOC and high bromide groundwater.

This study aims (1) to determine the seasonal and spatial distribution of THMs formed in chlorinated groundwater containing low levels of organic matter (0.4-0.8 mg L-1) and low to high levels of bromine (40-380 µg L-1), and (2) to evaluate the multi-route cancer risks associated with them. The study was conducted in Kayseri (Turkey), where drinking water is supplied from groundwater after chlorination only. THM formation in 50 water samples from 18 storage tanks and 32 distribution points was investigated to evaluate the spatial and temporal changes in THM concentrations for 12 months. The lifetime cancer risk associated with exposure to THMs through multiple pathways (i.e., oral ingestion, dermal absorption, and inhalation) was estimated for males and females. For a 12 month sampling period, the minimum and maximum THM concentrations varied from 2 µg L-1 to 17 µg L-1 and from 2 µg L-1 to 29 µg L-1 in storage tanks and distribution points, respectively. The ranges of median concentrations of THM were 5 µg L-1 to 9 µg L-1 in storage tanks and 5 µg L-1 to 12 µg L-1 in distribution points. In all samples dibromochloromethane was the dominant species, followed by bromoform, chloroform, and bromodichloromethane. The average values of total cancer risk associated with exposure to THMs via oral ingestion, dermal absorption, and inhalation for females and males were 1.31 × 10-5 and 1.25 × 10-5 in storage tanks, and 1.46 × 10-5 and 1.39 × 10-5 in distribution points, respectively. Although THM concentrations were very low, cancer risk values are 1.0 × 10-6 < CR < 1.0 × 10-4, which are higher than the negligible risk level (1.0 × 10-6).

Performance of ANN, SVM and MLH techniques for land use/cover change detection at Sultan Marshes wetland, Turkey.

Wetlands are among the most productive ecosystems that provide services ranging from flood control to climate change mitigation. Wetlands are also critical habitats for the survival of numerous plant and animal species. In this study, we used satellite remote sensing techniques for classification and change detection at an internationally important wetland (Ramsar Site) in Turkey. Sultan Marshes is located at the center of semi-arid Develi closed basin. The wetlands have undergone significant changes since the 1980s due to changes in water flow regimes, but changes in recent years have not been sufficiently explored yet. In this study, we focused on the changes from 2005 to 2012. Two multispectral ASTER images with spatial resolution of 15 m, acquired on June 11, 2005 and May 20, 2012, were used in the analyses. After geometric correction, the images were classified into four information classes, namely water, marsh, agriculture, and steppe. The applicability of three classification methods (i.e. maximum likelihood (MLH), multi-layer perceptron type artificial neural networks (ANN) and support vector machines (SVM)) was assessed. The differences in classification accuracies were evaluated by the McNemar's test. The changes in the Sultan Marshes were determined by the post classification comparison method using the most accurate classified images. The results showed that the highest overall accuracy in image classifications was achieved with the SVM method. It was observed that marshes and steppe areas decreased while water and agricultural areas expanded from 2005 to 2012. These changes could be the results of water transfers to the marshes from neighboring watershed.

Modelling the effects of meteorological parameters on water temperature using artificial neural networks.

Water temperature affects all biological and chemical processes in water; therefore, it is an extremely important water quality parameter. Meteorological factors are among the most important factors that affect water temperatures. The aim of this study is to develop an artificial neural network (ANN) model to investigate the effects of meteorological parameters on water temperatures at Kizilirmak River in Turkey. Water temperature data were collected from gauging stations on Kizilirmak River, and meteorological data were acquired from the nearest meteorological stations. Air temperature, wind speed, relative humidity, and previous water temperatures were formed the input parameters. The model output included water temperatures. All data were available for the 1995-2007 period, with occasional missing records. The activation functions of the ANN model and the number of neurons in the hidden layer were selected by trial-and-error method to find the best results. The root mean square error and the correlation coefficient between observed and simulated water temperatures were used to assess the model success. The best results were obtained by using sigmoid activation function and scaled conjugate gradient algorithm. This study showed that meteorological data can be used to simulate water temperature with ANN model for Kizilirmak River.

Optimization of solid content, carbon/nitrogen ratio and food/inoculum ratio for biogas production from food waste.

Biogas production from food waste has been used as an efficient waste treatment option for years. The methane yields from decomposition of waste are, however, highly variable under different operating conditions. In this study, a statistical experimental design method (Taguchi OA9) was implemented to investigate the effects of simultaneous variations of three parameters on methane production. The parameters investigated were solid content (SC), carbon/nitrogen ratio (C/N) and food/inoculum ratio (F/I). Two sets of experiments were conducted with nine anaerobic reactors operating under different conditions. Optimum conditions were determined using statistical analysis, such as analysis of variance (ANOVA). A confirmation experiment was carried out at optimum conditions to investigate the validity of the results. Statistical analysis showed that SC was the most important parameter for methane production with a 45% contribution, followed by F/I ratio with a 35% contribution. The optimum methane yield of 151 l kg-1 volatile solids (VS) was achieved after 24 days of digestion when SC was 4%, C/N was 28 and F/I were 0.3. The confirmation experiment provided a methane yield of 167 l kg-1 VS after 24 days. The analysis showed biogas production from food waste may be increased by optimization of operating conditions.

Use of Landsat land surface temperature and vegetation indices for monitoring drought in the Salt Lake Basin Area, Turkey.

The main purpose of this paper is to investigate multitemporal land surface temperature (LST) changes by using satellite remote sensing data. The study included a real-time field work performed during the overpass of Landsat-5 satellite on 21/08/2011 over Salt Lake, Turkey. Normalized vegetation index (NDVI), vegetation condition index (VCI), and temperature vegetation index (TVX) were used for evaluating drought impact over the region between 1984 and 2011. In the image processing step, geometric and radiometric correction procedures were conducted to make satellite remote sensing data comparable with in situ measurements carried out using thermal infrared thermometer supported by hand-held GPS. The results showed that real-time ground and satellite remote sensing data were in good agreement with correlation coefficient (R2) values of 0.90. The remotely sensed and treated satellite images and resulting thematic indices maps showed that dramatic land surface temperature changes occurred (about 2°C) in the Salt Lake Basin area during the 28-year period (1984-2011). Analysis of air temperature data also showed increases at a rate of 1.5-2°C during the same period. Intensification of irrigated agriculture particularly in the southern basin was also detected. The use of water supplies, especially groundwater, should be controlled considering particularly summer drought impacts on the basin.