International Journal of Water (20 papers in press)

Regular Issues

• Assessing irrigation network performance based on different climate change and water supply scenarios: a case study in Northern Iran  
  by Zohreh Dehghan, Farshad Fathian, Saeid Eslamian, Jan Franklin Adamowski 
  Abstract: The aim of this paper was to investigate the performance of irrigation networks under climate change, through a case study of a sprinkler irrigation network in Bilesavar, Northern Iran. The performance of the irrigation network was evaluated using WaterGems in terms of the equity and the adequacy of pressure, and minimum and maximum velocity at the outlets, based on different scenarios. The results showed that ETo may be around 6% higher than the baseline (1971-2000) by 2010-39, and 12% higher by 2050-79. Consequently, irrigation requirements may be higher in the future. Owing to climate change, it was seen that the irrigation network may experience challenges in terms of pressure and discharge supply. With increasing demand on the network, equity and adequacy indices of pressure distribution were seen to decline. In order to adjust to these changes, adaption strategies such as changes in the area of cultivation showed the greatest impact in reducing the volume and demand of water in the network. In general, the results showed that the various potential climate change secenarios may have a significant impact on irrigation network performance.
  Keywords: climate change; irrigation network; evaluation indices; vulnerability; Iran.
  
  
• Multiple regression modelling approach for rainfall prediction using large scale climate indices as potential predictors  
  by H. M. Rasel, M. A. Imteaz, F. Mekanik 
  Abstract: Some studies have established the associations with different climate indices (Southern Oscillation Index, Indian Ocean Dipole and Southern Annular Mode) and seasonal rainfalls of different parts of Australia. Nevertheless, maximum predictability of South Australian rainfall was only 20% with individual effects of potential predictor. To establish a better relationship for South Australian spring rainfall prediction, this paper presents two further investigations; a) a relationship of lagged climate indices with rainfall, and b) a combined influence of these lagged climate indicators on rainfall. Multiple linear regression (MLR) modelling was used to evaluate the influence of combined predictors. Three rainfall stations were selected from South Australia as a case study. It was revealed that significantly increased rainfall predictability has been achieved through MR models using the influences of combined lagged climate predictors. The rainfall predictability ranging from 41% to 45% has been achieved using combined lagged indices, whereas maximum 33% predictability can be achieved using individual climate index.
  Keywords: rainfall; ENSO; SAM; MR model; correlation; multicollinearity; forecasting.
  
  
• Introducing hydrochemical diagnostic tools to differentiate sources of salinity, southern Iran  
  by Mehdi Zarei 
  Abstract: Groundwater quality in southern Iran is typically degraded by a variety of salinity sources including seawater intrusion, saline lakes, salt diapirs, geothermal springs and groundwater evaporation. Twenty three sampling sites from the five major salinity sources of southern Iran were sampled in order to evaluate the skill of conventional hydrochemical methods, and to propose new diagnostic tools for differentiation of the salinity sources. The samples were analysed for major ions and minor constituents of lithium, bromide, strontium, boron and more than thirty trace elements. The results indicate that the conventional methods are not practical to differentiate salinity sources of southern Iran. Ion ratios of Na/Cl, Cl/Br, Li/Cl and B/Cl are more or less useful indicators for primary identification of some salinisation sources. However, the proposed composition binary diagrams of Br/Cl vs. Li/Cl, Sr/Cl vs. Li/Cl, and SO4/Cl vs. Li/Cl and ternary diagrams Li-Cl-Br, Sr-Li-Br and Cl-Li-B are the most feasible and confident tools to differentiate salinity sources of the study area. Finally, investigating the hydrochemical results of saline waters from different countries justifies that our proposed diagrams can be extensively used for differentiation of salinity sources with some calibration, especially for the district of evaporation.
  Keywords: salinity sources; diagnostic tools; salt diapir; geothermal; lake; southern Iran
  
  
• Rainwater tank analysis tools, climatic and spatial variability: a case study for Sydney  
  by Monzur Imteaz, Muhammad Moniruzzaman 
  Abstract: Among methods of analysing Rain Water Harvesting System (RWHS), a daily water balance method is most feasible and reasonably accurate. However, most of the relevant past studies using daily water balance method used continuous simulations using historical daily rainfall for a long period, and then provided an average expected water savings, which is the average of water savings from total simulation period. Through such averaged outcomes, rainwater tank users do not get an adequate insight of the expected realistic situation(s) in regards to variability of outcomes. With the impacts of climate change, such ranges of realistic outcomes are expected to be widening further. This paper presents analyses from a recently developed daily water balance model (eTank), which calculates rainwater tank outcomes under three different climatic conditions (i.e. dry, average and wet years). For comparison, eTank calculated results were compared with two other continuous simulation type water balance model, CSWBM and Raintank Analyser. It is found that CSWBM produced calculations water savings and reliabilities significantly vary with the eTank calculated water savings and reliabilities. Again, eTank calculated expected annual rainwater savings were compared with the expected water savings generated by widely used tool, Raintank Analyser, which uses historical daily rainfall data for many years and presents an average of all the calculated years cumulative water savings. It is found that Raintank Analyser calculated water savings closely match with the eTank calculated water savings in average year, which is reasonable and ascertains eTanks accuracy. Further, eTank was used to present expected climatic variations in water savings in Sydney, as well as expected spatial variations within the city.
  Keywords: rainwater tank; daily water balance, water savings, climatic conditions
  
  
• Analysis of drought transitions using log-linear models in Iran  
  by Mehdi Bahrami, Abdol Rassoul Zarei, Safie Chakav 
  Abstract: The prediction of drought category is performed using a log-linear model for three-dimensional contingency tables. The frequencies of drought categories are evaluated by a Standardized Precipitation Index (SPI) on a monthly time scale using the raw data obtained from seven stations in Iran. In this study, standardised precipitation index data of total period (60 years) sets were divided into three periods of 20 years and a log-linear model approach has been used to investigate differences relative to drought class (four drought severity classes) transitions among these three periods. The drought class transitions were measured for three periods to form a three-dimensional contingency table. According to the results, Mashhad, Tehran, Tabriz, Shiraz, Esfahan and Kerman stations had no significant changes in 95% level to number of drought classes, but in Zahedan station the drought severity had increasing trend. Results show that the drought behaviour in the second period is similar to that in the third. If just the second and third periods were compared it could be concluded that droughts were aggravating and this behaviour could be attributed to climate change. Supporting the common assumption, a trend for progressive aggravation of drought occurrence exists. Therefore, results are more consistent with the long-term natural periodicity; however, this hypothesis should be tested using longer time series.
  Keywords: log-linear models, drought, time series, three-dimensional
  
  
• Climate change fingerprints in the Lower Euphrates Basin: climate and flow data trend analysis  
  by Monzur Imteaz, Wesam Mahmood, Khaled Sagar, Abdullah Yilmaz 
  Abstract: It is expected that climate change will affect the global weather, snow coverage and ice melting, sea level rise, hydrological cycle, agriculture and forests, ecosystems and health. One of the most affected areas of climate change is hydrology and water resources. In addition to adverse impact of climate change, increasing uncontrolled abstractions from the upper basins is likely to worsen the situation in the lower basins. The Euphrates river, the longest in Western Asia, originates from the Armenian highlands and flows through Turkey, Syria and Iraq. The Lower Euphrates River Basin is predominantly in Iraq and is likely to be affected by climate change. Owing to the constant rise of water demands, to achieve a sustainable water management policy it is necessary to assess the impacts of climate change in the region. To assess trends of precipitation, temperature and streamflow to determine whether climate change impacts have already started at Lower Euphrates Basin, two popular non-parametric trend analysis methods, Mann-Kendal and Spearmans Rho tests, were applied in this study. Based on the time series data, a significant rising trend in temperature and a significant decreasing trend in streamflow were observed, which would be an important perception for policy making pertaining to water resources management in the Lower Euphrates Basin.
  Keywords: climate change, Euphrates river, trend analysis, temperature, precipitation, flow
  
  
• Thornthwaite-Holzman model for a wide range of daily evaporation rates  
  by Jaber Almedeij 
  Abstract: This study employs meteorological data from a weather station located in a coastal desert area in Kuwait, with a wide range of temperature, relative humidity, wind speed and evaporation measurements. The data have been used to test the well-known theoretical aerodynamic model of Thornthwaite-Holzman. The results show that the performance of this model is satisfactory for evaporation rates up to 10 mm/day, but there is an obvious systematic shift in accuracy beyond that from 10 to 40 mm/day. It is noted that the specific humidity difference proposed in the original model was assumed to have a linear correlation with evaporation rates. The study suggests modifying this assumption to become rather of a power form. The modification produced acceptable results based on subjective statistical criteria. This modification will help extending our ability for analysing hydrologic problems in different environments of high or low evaporation rates.
  Keywords: aerodynamic formula; specific humidity; coastal desert location; actual and potential evaporation; water cycle
  
  
• A groundwater ecosystem classification: the next steps  
  by Peter Serov, Laura Kuginis 
  Abstract: Groundwater performs a far more complex role in the landscape than previously considered by supporting a vast array of organisms and ecosystems. These ecosystems have a range of values and functions including use as bioindicators of groundwater health, adding a substantial biodiversity value to a region and contributing to the overall ecological function of a resource on which much of the worlds human population are dependent. Ecosystems dependent on groundwater were first recognised and classified in Australia in 1998. Since then, the number of recognised types of Groundwater Dependent Ecosystems or GDEs has greatly increased to the point where they are now known by many names and occur in almost every environment around the world. The lack of a standardised classification system for GDEs, however, has caused some difficulty in grouping them appropriately and managing the diverse range of ecosystems in a logical and meaningful manner. An ecohydrogeological classification system for GDEs is presented here that was developed to protect high value water dependent ecosystems including groundwater dependent ecosystems. The need for a new classification system for this specialised group of ecosystems became apparent when the existing classifications could not adequately encompass, characterise, order, or prioritise the large number of known GDEs. In this paper we review the previous classification systems that have included groundwater ecosystems worldwide and discuss the attributes and ecological drivers used. This paper focuses particularly on GDEs occurring in Australia. A new, more comprehensive classification scheme for aquifers is also presented based on structure, lithology and degree of confinement.
  Keywords: Groundwater dependent ecosystems; GDE; classification; wetlands; phreatophytes; submarine groundwater discharge; stygofauna; caves; karsts; hyporheic; base flow; springs.
  
  
• Application of artificial neural networks to predict peak flow of Surma River in Sylhet zone of Bangladesh  
  by Abul Abrar Masrur Ahmed, Syed Mustakim Ali Shah 
  Abstract: River flow analysis and prediction is an important task in water resources planning particularly for a disaster prone agricultural country like Bangladesh. Present study used two ANN models namely radial basis function (RBF) and multi-layer perceptron (MLP) to analyse Surma River flow and estimate its peak flow concentration based on five input parameters. The performances of selected models were measured using the correlation coefficient (R), mean absolute error (MAE) and model efficiency (EFF%). However, RBF network model performed better than MLP network model with high model efficiency (99.55%), low mean squared errors (38.60) and high correlation coefficient (0.996), where the optimum number of neuron was 18 for RBF and 22 for MPL network. Moreover, proposed ANN models could be used successfully in estimating the peak-flow of the Surma River which would facilitate water resources management policy of this region.
  Keywords: artificial neural network; radial basis function; multilayer perceptron; Surma River; peak flow; Sylhet; Bangladesh.
  
  
• Intricacies, challenges and implications: the governance of Tadlac Lake, Los Banos, Laguna, Philippines  
  by Bing Baltazar Brillo 
  Abstract: Grounded on the lacuna in literature the scarcity of scholarly works on lake govern-ance and small lakes in the country, and the notable transformation of Tadlac Lake from a threatened and poorly regulated lake to a potential model for governance and ecotourism development, this article documents the small lake and examines its governance specifics, challenges and implications. It argues that the governance of Tadlac Lake can be characterised as intricate, watershed-based, hierarchical, participatory and centralised. It also contends that the lakes development its transformation from an aquaculture-based lake into an ecotourism-oriented lake illustrates a key lesson for developing small lakes in the country. To close, the article hopes to instigate more governance studies on lakes, particularly small lakes, owing to their number (in the country and globally) and the fact that in the country, many are surrounded by impoverished communities.
  Keywords: development; governance; Philippines; lake; small lake; Tadlac Lake.
  
  
• Variation in rainfall trend at upstream: a threat to the filling schedule of Hirakud Reservoir,India  
  by Krishna Kumar Gupta, Anil Kar, Joygopal Jena, DiptiRanjan Jena 
  Abstract: Hirakud is one of the major reservoirs of Mahanadi basin of India. The reservoir drains water from a catchment of 83,400 sq km out of the total catchment of 141,569 sq km of the basin. The filling of the reservoir is very important as it fulfills a number of demands, and in extreme rainfall events it plays an important role in flood control. This study focuses on the possibility of trend in monthly rainfall of upstream districts of the reservoir by applying the Mann-Kendall test. It was found that out of 23 districts of upstream Hirakud, no trend was found for June (except Bilaspur showing rising trend) and October rainfall in any of the districts, two districts namely Kwardha and Surguja show falling trend in the month of July and the rest show no trend, in August the same Kwardha and Surguja show falling trend, and Nawarangpur, Gadachiroli and Bastar show rising trend. Almost 19 districts show falling trend in September at 5% significance level. However, the reservoir showed declining trend during August and no trend in rest of the monsoon months. As September is sensitive towards the filling of the reservoir, as well as for flood control, the falling trends of rainfall will definitely alert the planners. The effect of climate change as well as changes in landuse and storing/releases of upstream reservoir are suspected to be to blame for the irregularities found in the pattern. Integrated basin management is also necessary for smooth operations of all the reservoirs.
  Keywords: trend; Mahanadi; Hirakud; Kendall rank; climate change.
  
  
• Modelling and sensitivity analysis of river flow in the Upper Indus Basin, Pakistan  
  by Firdos Khan 
  Abstract: Undoubtedly, it is important to model the average and extreme phenomena in earth sciences disciplines such as hydrology under uncertain and changing climate conditions. The issues become more important when we deal with reservoir management, flood forecasting and irrigation. In this paper, we model the average and extreme river flow in the Indus River at the Upper Indus Basin upstream of Tarbela Reservoir. For modelling the average river flow, we used the popular classes of time series models including the Autoregressive Integrated Moving Average (ARIMA) and Autoregressive Conditional Heteroscedasticity (ARCH) models. For modelling the extremes, statistical probability distributions dealing with extremes in the tails of the distributions of considered phenomena were used. Starting with different models and distributions we finally choose the one which performs best among the competing models and distributions, respectively. Preference is given to positively skewed and extreme value distributions. For model selection, we incorporated both model selection criteria and forecast performance evaluation statistics. Whereas in selecting the best distribution, goodness-of-fit statistics, model selection criteria and post-fit simulations from each distribution have been used. Finally, when modelling extremes we noted that different probability distributions may be used for the same data, depending on whether interest is in lower or higher order moments.
  Keywords: hydrological extremes; river flow; extreme value distributions; Tarbela Reservoir; Upper Indus Basin; time series models.
  
  
• Precipitation change assessment over upper Bagmati river basin using regional bias corrected GCM data  
  by Binaya Kumar Mishra, Srikantha Herath 
  Abstract: Climate change impacts on precipitation patterns have been assessed over the upper Bagmati river basin, Nepal. General circulation model (GCM) projections, which are widely used to assess climate change impacts, consist of significant biases due to oversimplification of the global climate system. Bias in daily GCM precipitation was corrected at grid cell scale using regional quantile-based bias correction technique. In this technique, the bias correction pattern is established by comparing GCM grid data with nearby observation station data. The concept of homogeneous precipitation regions was used to transfer statistical characteristics of nearby observation stations to ungauged cells. Calibration and validation of the regional bias correction technique was performed over 1979-2003. Comparative precipitation change assessments were carried out at annual, monthly and daily time scales, considering bias corrected precipitation data of 1979-2003 and 2075-2099 periods, respectively.
  Keywords: bias correction; climate change; GCM; homogeneous precipitation regions; impact assessment.
  
  
• Spatio-temporal analysis of groundwater level in an arid area  
  by Abbas Alipour, Seyedmostafa Hashemi, Sajad Bagheri Said Shokri, Mojtaba Moravej 
  Abstract: Sustainable management of precious groundwater resources is vital to growth, especially in arid areas that heavily depend on it. In this study, monthly groundwater level time series of 322 observation wells in Sistan and Balouchestan province of Iran in the period of 2002 to 2012 were used as a measure of sustainable development. Results show that 95.6% of the area of the province's aquifers are non-stationary. In addition, 5.5%, 13.7% and 80.8% of studied wells show no increasing or decreasing trend at 5% significance level. Spatial analysis reveals that the groundwater level of 7.3%, 11.3% and 81.6% area of groundwater resources have increased, sustained and decreased, respectively. According to the calculated slope of trends, the province experienced -69.72 (MCM) decline in groundwater resources per year in the 10 years studied. It means that sustainable development criteria were totally neglected. Increasing water use efficiency in the agriculture sector by 6% and water reuse by 78% will solve the issue.
  Keywords: ADF test; aquifer; groundwater level; Şen innovative trend analysis method; sustainable development; Sistan and Balouchestan; stationary tests; water efficiency; water reuse.
  
  
• Comparison of two statistical climate downscaling models: a case study in the Beijing region, China  
  by Binbin Guo 
  Abstract: Climate change can have a significant impact on the hydrological cycle. Temperature, precipitation and evapotranspiration are important climate factors, whose future trend predictions have important significance for assessing the impact of future climate change. This article focuses on the comparison of two climate statistical downscaling models applied to global climate model (GCM) predictions for the Beijing region. The GCM predictions are for the high greenhouse gas emission scenarios (A2 scenarios) and the low greenhouse gas emission scenarios (B2 scenarios). The two statistical downscaling models used are the Statistical DownScaling Model (SDSM) and the Automated Statistical Downscaling Model (ASD). Through the analysis of the evaluation indices in the calibration and validation periods, the results show that both downscaling models simulate the temperature and evapotranspiration well, but the simulation of precipitation is not as good as that of other climate factors. The overall performance of ASD model is slightly superior to that of SDSM model, especially in the process of predictor's selection. The future climate change downscaled by the two models shows an analogous trend as well. The temperature and evapotranspiration show a general increasing trend. The precipitation shows a different trend with an increasing trend in the south and a decreasing trend in the north.
  Keywords: HadCM3; SDSM; ASD; climate change; Beijing area.
  
  
• Hydraulic interconnections study of Seropan-Ngreneng-Bribin underground rivers in Gunung Kidul karst area , Indonesia, using a tracer technique  
  by Paston Sidauruk, Rasi Prasetio, Satrio Satrio 
  Abstract: A series of studies have been conducted to investigate the hydraulic connection between Seropan underground rivers with other underground rivers in Gunung Kidul karst area. Stable isotopes composition and hydrochemical contents of collected samples from several sources around Seropan cave, together with tracer tests, have been conducted in the area to help to establish hydraulic connection of Seropan underground river with adjacent underground rivers. The tracer tests revealed that subsurface water flowed from Seropan to Ngreneng cave with average travel time about 6 hours and 13 minutes. Analyses of breakthrough curves found from the tracer tests hinted that the flow is a direct flow with a large well-developed conduit without any significant delay due to water depressions and dead zones. Although downstream flow paths of Seropan underground river were successfully determined, upstream hydraulic connection were not established. Future study should focus on the upper part (upstream) of Seropan cave.
  Keywords: karst area; stable isotopes; hydrochemistry; tracer test; hydraulic interconnection; underground river; caves.
  
  
• Evaluating water productivity of tomato, pepper and Swiss chard under clay pot and furrow irrigation technologies in semi-arid areas of northern Ethiopia.  
  by Amanuel A. Gebru, A. Araya, Solomon Habtu, Tsegay Wolde-Georgis, Daniel Teka, Luceita G. Martorano 
  Abstract: Managing irrigation water is among the critical issues to address food insecurity under climate change and variability conditions. Irrigation is suggested as one of the adaptation practices commonly implemented to reduce climate-related risks. However, there is a scarcity of water in many drylands, and identifying an efficient and effective irrigation system is crucial. A comparative study was undertaken between bar-shaped clay pots and furrow irrigation on tomato, pepper and Swiss chard crops in northern Ethiopia during the cropping season of 2014/2015. Results were compared on the basis of yield, water productivity and economic performance. The yields of Swiss chard, tomato and pepper were increased by up to 51%, 32% and 30%, respectively, in the bar-shaped clay pot irrigation system compared with the control. Water saving was also considerably increased by 40.6%, 41.2% and 41.7% for the respective crops compared with the control. Similarly, the water productivity of Swiss chard, tomato and pepper was 10.9, 4.2, and 1.8 kg m-3, respectively. Further research on the suitability of bar-shaped clay pot irrigation on various soils and crops is recommended.
  Keywords: bar-shaped clay pot; furrow irrigation; water productivity; yield response.
  
  
• Status of domestic water supply and prospects of rainwater harvesting in south-eastern Nigeria  
  by Chidozie Nnaji, Chijioke Edeh, John Nnam 
  Abstract: Two major cities (Enugu and Abakaliki) in south-eastern Nigeria were investigated with a view to ascertaining current rainwater harvesting (RWH) practices and prospects for supplementing available supply. The methods employed in this study include: distribution of questionnaires for the determination of water consumption and supply specifics, and rainwater harvesting practices; and use of water balance and optimisation to maximise water storage for dry season supply. Using the income level supplied by respondents, data were categorised into low income, middle income and high income, and analyses were performed along these income delineations. For Enugu, the per capita water consumption is 23.7l pcd, 34.45l pcd, and 67.05l pcd for the low income, middle income and high income groups, respectively. For Abakaliki, the per capita water consumption is 15.56l pcd, 28.08l pcd, and 50l pcd for low income, middle income and high income groups, respectively. In Enugu, 47%, 61.8% and 37.9% of the low income, middle income and high income groups, respectively, practice RWH. In Abakaliki, the corresponding proportions are 67.2%, 48.8% and 46.9%, respectively. Rainwater can meet 100% of the current water consumption of low income groups of both cities for bungalows and up to five storey buildings or twelve flats housing 72 residents. For buildings taller than five storeys housing the middle and high income groups, the RWH potential can be improved by adopting integrated rainwater harvesting (IRWH) where spill from one building is used to make up for deficit in another.
  Keywords: rainwater harvesting; rainwater harvesting potential; water consumption; optimisation.
  
  
• Urbanisation and hydraulic geometry response: a model approach  
  by Adeyemi Olusola, Toyin Fashae 
  Abstract: As the process of urbanisation continues within a river basin, the response of the river channels within the basin alters, resulting in a variety of responses from the river channel: increase in the area of low infiltration capacity and the efficiency or speed of water transmission in channels or conduits, variations in peakflow discharges and stream velocity along a channel, etc. This study pursues further to understand the impact of urbanization on channel geometry of the Ona River Basin using the Hydrological Modelling System (HEC-HMS). Additionally, the study seeks to provide answers on the downstream hydraulic geometry of the main segment of the Ona River as it relates to downstream discharge. The hydrograph of the basin shows that the present discharge of the channel from this study is 5177.32 cubic metres per second with a peak time of 10.52 min, while the moderated channel response due to increase in the level of urbanisation discharges 5342.19 cubic metres per second and peaks at 10.45 min. This shows an increase of about 3% in discharge. This study has been able to show that the rate at which urbanisation contributes to an increase in discharge of urban rivers is not high. This suggests that urban flooding is attributable to many factors, one of which is the present level of urbanisation and which may not necessarily carry a very high weight. The exponents of the hydraulic geometry of the river, 0.15, 0.28, 0.57 (width, depth and velocity, respectively) indicate the variability within these three variables and supports the fact that f and m increase while b decreases for cross-sections studied within the basin. The implication of DHG suggests that with increasing bankfull discharge, the bankfull variables of mean depth, mean width and mean velocity also increase systematically downstream.
  Keywords: Ona River; HEC-HMS; hydrograph; bankfull discharge; urbanisation.
  
  
• Modelling non-stationary extreme streamflow in peninsular Malaysia  
  by Nur Amalina Mat Jan, Ani Shabri, Jean Hounkpe, Basri Badyalina 
  Abstract: Global change and global warming have raised concerns among hydrologists about the use of the stationary assumption (independent and identically distributed flood series) in infrastructure-designed methods. This confirms the necessity of evaluating the stationary or non-stationary behaviour of hydrological variables before deriving designed flood for infrastructure projects and flood mitigation. Trends were evaluated in the annual maximal streamflow of 49 stations in peninsular Malaysia, over 50 years (1960-2009) using the Mann-Kendall and Spearman Rho trend tests. Three models, a stationary model (GEV0), a non-stationary model with location parameter linear function of time (GEV1) and a non-stationary model with location and log-transformed scale parameters as a linear function of time (GEV2), were considered for stations with a significant trend. The likelihood ratio test, the Akaike information criteria and the Bayesian information criteria were used to determine the best-fitting model. It was found that a quarter of the analysed stations show statistically significant trends in their annual maximal streamflow. Among the 14 stations showing statistically significant trend at 5% level, 10 stations were found to better fit with the non-stationary models (GEV1 and GEV2) based on three goodness of fit measures. These results indicate the importance of taking into consideration the non-stationary behaviour of the flood series in order to improve the quality of flood estimation.
  Keywords: stationary; non-stationary; streamflow; flood frequency analysis; generalised extreme value distribution; Mann-Kendall and Spearman Rho trend test.