Introduction

Global warming due to the Green House Gas (GHG) concentration increases leading to climate change tendencies, which attracted researchers’ and people’s attention due to the unexpected and adverse results on sectoral damages (Hobbs 2003). According to the observed global average temperature trends in previous records, the global average temperature has increased by about 0.74 °C since comparing pre-industrial time (UNFCCC 2007). In the same period, although the precipitation on the land surface has decreased in the tropics, increases are observed especially in the middle and high latitudes (IPCC 2007). These variabilities are often and intensively beginning to occur as extreme precipitation and high summer temperatures (Dabanli 2019). Recent climate modeling studies are conducted with identification of these change effects and gained an improved ability to simulate many aspects of climate change, variability, and extreme events (IPCC 2014).

In climate change and hydrology studies, rainfall and temperature parameters are accepted as the major factors underlying unexpected adverse effects on nature, agriculture, and society (Cheung et al. 2008). As told by Riddle and Cook (2008), these parameters have crucial importance for spatiotemporal analysis of climate change to sustain logical analysis output. Mathematical model for climate change impacts and predictions is helpful to prevent adverse impacts of extreme rainfall and high temperatures (Tabari and Talaee 2011). Trend analysis for water resources has significant importance over drought, water scarcity, and extreme precipitations, which may result in flooding. In the northern Mediterranean region, there are frequent water-stressed basins due to climate change impacts. Vulnerable populations to unexpected impacts on these regions are estimated to be from 1.4 to 2.1 billion (Arnell 2004; Alcamo et al. 2006).

While total precipitation amount increases, precipitation variability has been changed in mid-latitudes. Frequency of intensive precipitations tends to increase in those regions (Dabanli 2019). In particular, Turkey has witnessed several flooding on Black sea regions due to intensive rainfall. In July 2019, more than six people disappeared after flooding in Duzce province in the northern region. Prevention against flood disaster, riverbed restoration, and watershed planning are essential. In this planning, rainfall and temperature trend magnitude should be considered while calculating risk and other precautions. In most cases, observed data is evaluated by calculating the average values by ignoring the extreme events. Consideration of extreme data instead of average values may help to decide on qualitative scientific definitions and predictions in factual practices (Şen 2010; Dabanli et al. 2017).

Similarly, temperature records are expected to increase, while monthly average precipitation records decrease in mid-latitudes (Gocic and Trajkovic 2013). Jiang et al. (2014) revealed that precipitation decrease due to climate change may also reduce the flow rates around 20%. The limitations in the spatial and temporal scopes of the monitoring networks cause uncertainty in hydrologic variables’ trends (Huntington 2006). Trends of monthly and annual record values have been examined by many researchers and organizations depending on various parameters and different methods (Kottegoda 2015; Hirsch and Slack 1984; Loftis et al. 1991; Lettenmaier et al. 1994; Dabanli et al. 2017; Mortsch et al. 2009; Zhang et al. 2001; Helsel and Hirsch 2002; Libiseller and Grimvall 2002; Önöz and Bayazıt 2003; Kahya and Kalaycı 2004; IPCC 2007; Pujol et al. 2007; Kumar et al. 2010; Tabari and Talaee 2011; UNDP 2012; Şen 2013; Sonali and Nagesh Kumar 2013).

In this study, meteorological variables (temperature and precipitation) are analyzed by Şen’s 1:1 (45°) innovative trend assessment (ITA) method by consideration of data from the Yesilirmak basin (based on historical observation of 1970–2013 years) in the north part of Turkey. Subsequently, climate change impacts on meteorological variables are also interpreted under the light of the same approach.

Study area and data

Yesilirmak river basin is located in the northern part of Turkey with approximately 39,595 km2 area. The basin is covered by Eastern Black Sea, Coruh, Firat, and Kizilirmak basins as presented in Fig. 1. The Yesilirmak river (519 km length) originates from Köse mountains to the southwest of Sivas province. From the upstream to the downstream, it passes through the plains of Tokat, Turhal, and reaches to Amasya and Carsamba plains. The Yesilirmak basin covers a part of Tokat, Samsun, Amasya, Çorum, Sivas, Yozgat, Gümüşhane, Giresun, Erzincan, Ordu, and Bayburt provinces with the main branches of the river as Kelkit, Cekerek, and Tersakan Streams. The Yesilirmak basin is divided into 32 sub-basins by the General Directorate of State Hydraulic Works (DSI). Temperature and precipitation data obtained from General Directorate of Meteorology (MGM) are given in Table 1. In the basin, average annual precipitation is 646 mm and average annual flow is 5.80 km3/year for Yesilirmak river.

Fig. 1
figure 1

Study area and observation stations in Yesilirmak basin

Full size image
Table 1 Observation stations and related information in Yesilirmak basin
Full size table

Methodology

Yesilirmak basin is selected for the trend application studies due to its utmost importance for water resource sustainability and management studies. For this purpose, five regularly measured temperature and precipitation data at different sites are selected to show the present situation.

A recorded hydro-climatic time series data is divided into two equal length groups from the beginning date to the end and both sub-series are separately sorted into ascending order. The first sub-series (Xi) is located on the X-axis and the other sub-series (Xj) on the Y-axis as seen in Fig. 2. If data plots appear on the 1:1 (45°) straight line or near to it, it implies that there is no trend (Dabanlı et al. 2016). It is obvious that in the case of increasing (decreasing) monotonic trend, the scatter points fall above (below) the 1∶1 straight line.

Fig. 2
figure 2

Illustration of decreasing, increasing, and trendless regions (Dabanli et al. 2016)

Full size image

In this study, trend envelope lines are applied for significant assessments of possible trends. These envelope lines are generated as parallels to 1:1 (45°) straight line. Thus, numerical trend assessments can be achieved easily by consideration of ± %5 or ± %10 trend envelopes.

Result and discussion

Six observation stations are selected for the trend analysis study. Figure 3 shows the results for the precipitation and temperature trend and anomalies for Susehri station. It is possible to see that the precipitation data have a significant decreasing trend. This figure also illustrates that temperature trend has increasing character, especially at the medium range, where average precipitation values have a − 4% decrease. In the same figure, temperature records have ascending trends (5–10%). In particular, one can see those medium temperatures between 0 and 10 °C, which are expected to increase (+ 7%) more rapidly in future. All trend slopes are illustrated in the figures and numerical values are presented in Table 2. Precipitation and temperature anomaly limits are detected as ± 14 mm and ± 3 °C, respectively.

Fig. 3
figure 3

Precipitation and temperature trends and anomalies in the Susehri station Figures 3–7 contain poor-quality text inside the artwork. Please do not re-use the file that we have rejected or attempt to increase its resolution and re-save. It is originally poor; therefore, increasing the resolution will not solve the quality problem. We suggest that you provide us the original format. We prefer replacement figures containing vector/editable objects rather than embedded images. Preferred file formats are eps, ai, tiff, and pdf. Figure 2 contains text below the minimum required font size of 6pts inside the artwork, and there is no sufficient space available for the text to be enlarged. Please provide replacement figure file.Please see attachment. All figures have been changed in high quality and saved as PDF file.

Full size image
Table 2 Trend analysis results summary
Full size table

Tokat is located in the southeastern cost of the Yesilirmak sub-basin. Figure 4 shows that precipitation innovative template scatters have a descending trend for all range data values. In precipitation, − 4% decreasing trend is detected for low and medium ranges. In high ranges, this decreasing trend is calculated as − 3%. Temperature observations at this station have a significant increase in all ranges. They are calculated as + 5%, + 4%, and + 7% for low, medium, and high ranges, respectively. Precipitation and temperature anomaly limits are detected as ± 12 mm and ± 2.5 °C, respectively.

Fig. 4
figure 4

Precipitation and temperature trends and anomalies in the Tokat station

Full size image

Dokmetepe is located in the southeastern part of the Yesilirmak sub-basin. Figure 5 shows that precipitation innovative template scatters have a descending trend for all data range. In precipitation, − 4% decreasing trend is detected for low and medium ranges. In high ranges, this decreasing trend is calculated as − 3%. Temperature observations at this station have significant increase in all sub-groups. These rates are calculated as + 3%, + 4%, and + 7% for low, medium, and high ranges, respectively. Precipitation and temperature anomaly limits are ± 17 mm and ± 2 °C, respectively.

Fig. 5
figure 5

Precipitation and temperature trends and anomalies in the Dokmetepe station

Full size image

Corum station trend templates for precipitation and temperatures are presented in Fig. 6, where precipitation scatters have descending trend (− 3%) for all data ranges with the exception that 2 data records are out of this behavior and they can be treated as possible extreme values. Temperature records at this station have significant increase around + 5% trend envelope lines. Especially, around 10 °C, + 10% strong trend behavior is seen. Average trend slopes are calculated for low, medium, and high ranges as + 5%, + 4%, and + 7%, respectively. Precipitation and temperature anomaly limits are detected as ± 15 mm and ± 2.5 °C, respectively.

Fig. 6
figure 6

Precipitation and temperature trends and anomalies in the Corum station

Full size image

Havza is located in the northwestern cost of the Yeşilırmak basin. Figure 7 indicates that precipitation innovative trend template scatters have significant decrease as presented for precipitation at the Tokat station. According to figure, precipitation records can be split up with 80 mm. Besides, temperature records at this station have significant increase in all ranges. Strong consistency of trend is easily detectable between the Havza and Tokat stations. The precipitation and temperature characteristics are almost similar for these two stations. Precipitation and temperature anomaly limits are detected as ± 25 mm and ± 2 °C, respectively.

Fig. 7
figure 7

Precipitation and temperature trends and anomalies in the Havza station

Full size image

In precipitation, low, medium, and high ranges are considered at low levels than 40 mm, at a medium level between 40 and 80 mm, and at high level ranging more than 80 mm, respectively. These limits are determined based on expert judgment considering average values. Detected trend slopes are presented in Table 2 based on trend envelope lines. In this table, significant decreasing trends are valid for precipitation in all ranges. Results prove that although basin has high precipitation above country average, precipitation trends tend to decrease. Precipitation anomalies are fluctuating around 15–20 mm average in basin level.

In a similar manner, temperature sub-group limits are determined as < 0 °C, 0–10 °C, and < 10 °C for low, medium, and high ranges, respectively. From the graphs, temperature anomalies are fluctuating on the average around 2 °C in basin level. Increasing trends are calculated for temperature records. These outputs strengthen climate change expected in the Yesilirmak basin by increasing temperatures.

Conclusions

In this study, the innovative trend analysis (ITA) method is employed for meteorological time series trend detections. The ITA approach provides categorical trends within the low, medium, and high data ranges, whereas other popular trend methods do not provide categorical trends, but single monotonic trend component within the time series. For the implementation, monthly precipitation and temperature records are considered or the 44-year duration from Yesilirmak basin for trend analysis and anomaly presentation. It is observed that the ITA trend test provides significantly decreasing trend tendencies for precipitation in low, medium, and high levels on an average − 3.4%, − 3.8%, and − 2.4%, respectively. In contrast, precipitation trends are consistent with global climate change expectations and, temperature records have significantly increasing trends. In sub-levels of temperature, this strong trend tendencies are detected as + 4.6%, + 4.8%, and + 7.2% from low to high ranges, respectively. Moreover, the trends can be detected at different categories with consideration of the trend envelope lines. Combination of these envelope lines with the existing ITA method provides a supportive analysis way for trend identification categorically. It also provides reliable interpretations for drought (low) and flood (high) values to manage climate change effects on water resources in future studies.