Climate variability, trend, and associated risks: Tana sub-basin, Ethiopia

Climate change is mainly referred to as the long-term fluctuations in weather parameters of a large area with statistical significance (Getachew & Manjunatha 2021). Climate change is one of the world's major challenges in the 21st century (Field 2014; Abidoye & Odusola 2015; Reidmiller et al. 2018), and its adverse impacts challenge people's socioeconomic activities, livelihood, health, and food security (Clarke et al. 2012). Global warming and climate variability are the emerging foremost environmental problems and global threats in the 21st century, particularly in developing countries (Birara et al. 2018; Habte et al. 2021). Climate variability and change, its impacts, and the associated vulnerabilities, are rising environmental issues worldwide (Makenzi et al. 2013). The fifth assessment report of the IPCC (IPCC 2013) indicated that the global mean temperature showed a warming trend of 0.85 °C (0.65–1.06) over the period 1880–2012. Already, most regional studies use long-term changes in temperature and rainfall patterns as a proxy indicator of climate change (Enyew 2014; Addisu et al. 2015; Birara et al. 2018, 2020; Berihun et al. 2019; Esayas et al. 2019; Tenagashaw & Andualem 2022). Many of the developing countries, mainly those found in sub-Saharan Africa, are significantly influenced by the global average temperature rise and its consequences (IPCC 2012). The studies by Belay et al. (2021) and Getachew & Manjunatha (2021) indicated that Earth's climate change and variability are the results of either natural variability or anthropogenic (human) activity, in combination. The anthropogenic causes arise from various activities such as excessive usage of fossil fuels, deforestation, and changes in agricultural practices are believed to increase the atmospheric concentration of greenhouse gases and temperature (UNFCCC 2007; Saroar & Filho 2016), and could alter the natural atmospheric processes for many decades (Hassan et al. 2014), while the natural causes include changes in solar activities, orbital parameters, and volcanic eruptions. Anthropogenic (human) activity has changed the earth's environment over the past century, while more changes are expected in the coming few decades, even if strong mitigation measures are taken (IPCC 2013).

The countries located in the sub-Sahara region include Ethiopia, whose economy is significantly affected by climate variability (Birara et al. 2018). The Blue Nile River Basin is one of the region's most sensitive basins to changing climate and water resource variability (Kim & Kaluarachchi 2009). Various studies have investigated historical trends of climate change and variability in Ethiopia. For instance, a 0.37 °C per decade increase was observed in minimum temperature between 1951 and 2006 by McSweeney et al. (2008), whereas, a 0.2–0.28 °C rise per decade in the average annual maximum temperature between 1960 and 2006 was observed by Keller (2009) and Eshetu et al. (2014). El Niño/Southern Oscillation has a significant influence on rainfall over Eastern Africa, especially during the October–December rainfall season (Awange et al. 2014; Omondi et al. 2014). The studies by Abera et al. (2020) indicated that high rainfall variability in Ethiopia is observed especially in the area where agricultural-dependent rural people are densely populated (highland regions). The study area is one of the areas in Ethiopia influenced by population pressure dominantly dependent on agriculture (Abera et al. 2020) and high expansion of cultivation practices (Tesfaw et al. 2023). A rapidly growing and dense population in the area is putting unprecedented pressure on natural resources. The annual and seasonal variability of rainfall significantly affected agricultural productivity, pastoralists, and animals in Ethiopia (Seleshi & Zanke 2004; Makenzi et al. 2013). Lake Tana is a lake in the Tana sub-basin, which is highly sensitive to variations in rainfall as well as variations in river inflows and evaporation (Kebede et al. 2006). Setegn et al. (2009) showed that inflow river discharge to Lake Tana contributes over 90% of the lake inflow. It is, thus, very likely that changes in river inflow would also change the volume of the lake and the water balance, which could ultimately adversely impact the lake ecosystem. Analyzing the spatial-temporal distribution of rainfall and detecting trends is the key to healthy ecosystem functions and sustaining agricultural production (Krishan et al. 2012; Meshesha et al. 2018; Worku et al. 2018). Therefore, this study was initiated to analyze climate variability, trends, and associated risks of the climate parameters (temperature and rainfall) in the Tana sub-basin, Ethiopia. The seasonal and annual variability of rainfall and temperature were assessed, and the concentration of precipitation was computed. The findings of this study will provide valuable information for planners and decision-makers in the region. Understanding the variability and trends of temperature and rainfall is crucial for developing appropriate adaptation and mitigation strategies to address the potential risks associated with climate change. This information can also be used to inform agricultural practices, water resource management, and infrastructure development in the Tana sub-basin. Overall, this study will contribute to a better understanding of the climate dynamics in the Tana sub-basin and provide valuable insights for sustainable development and climate resilience in the region.

According to the annual rainfall characteristics of Ethiopia, there are three rainy seasons (Gissila et al. 2004; Segele & Lamb 2005; Korecha & Barnston 2007; Haile et al. 2009). The country's main rainy season is from June to September, the dry season from October to January, and the small rainy season from February to May. Unimodal rainfall distribution occurs in the study area (Getachew & Manjunatha 2021), while its mean annual rainfall is estimated to be 1,280 mm (Setegn et al. 2008; Birara et al. 2018). Most rainfall falls during the major rainfall season from June to September (Alemu et al. 2020). The mean annual temperature of the study area is 21 °C by Weldegerima et al. (2018). The annual actual evapotranspiration of the sub-basin was 1,036 mm by Allam et al. (2016) and 733 mm by Setegn et al. (2008).

The climatic variables used for the study, including rainfall (1981–2020) and maximum and minimum temperature (1981–2016), were obtained from the Enhancing National Climate Services (ENACTS) products provided by the National Meteorological Agency (NMA). The ENACTS products (https://iri.columbia.edu/resources/enacts/) used in this study were derived from a grid and blended dataset, which combined ground observation station data with satellite estimates for rainfall, digital elevation models, and reanalysis products for temperature (Dinku et al. 2017).

The process involved applying the average weight method to generate time series climate data from the ENACTS grid product for the ten regions of the study area. The quality of the data was ensured by checking it against ground observation station data. In addition, the ENACTS dataset, along with nearby observed data, underwent performance evaluation using criteria such as Nash–Sutcliffe efficiency, coefficient of determination (R²), and Percent Bias. This comprehensive approach to data collection and validation ensures that the study is based on high-quality and reliable climate data, which is essential for accurate and meaningful analysis.

The coefficient of variation (CV), seasonality index (SI), precipitation concentration index (PCI), Mann–Kendall (MK) trend test, and Sen's slope estimator were used to determine the variability, trend, duration, and magnitude of annual, seasonal, and monthly precipitation and temperature in the study area (Verma et al. 2022; Kumar et al. 2023). Monthly rainfall and temperature data were used. Monthly rainfall and temperature data were used covering the period 1981–2020 and 1981–2016, respectively, as an input.

The value of the SI varies from 0 (when all months share the same amount of rainfall) to 1.83 (when all rainfall incidences occur in a single month). Walsh & Lawler (1981) also proposed rainfall regime classification based on SI values (Table 1). The classification is further redefined based on the rainfall duration during the year (Table 1).

The value of PCI less than 10 indicates the uniform distribution of precipitation; values between 11 and 15 represent moderate precipitation concentration; values between 16 and 20 indicate the irregular distribution of precipitation and the value which is above 20 unit shows a strong irregular precipitation distribution across the area (Oliver 1980; Belay et al. 2021; Edo et al. 2021).

Then, the MK test from the Z value was computed based on monthly, seasonal, and annual rainfall time series data.

The temporal variation of monthly rainfall was high compared to the temporal variation of monthly temperature in the study area, as indicated in Figures 2, 4, and 5. The significant variability of monthly rainfall across the region emphasizes the dynamic nature of precipitation patterns in the study area. This variability likely has important implications for various sectors, including agriculture, water resource management, and ecosystem dynamics (Wubneh et al. 2023). High rainfall variability (CV > 30) was observed from February to May, October to December, and in January for some parts of the regions (30.99% of the sub-basin) while in January (61.36%), June (15.3%), July (20.23%), and September (62.51%) of some regions show moderate rainfall variability (20 < CV > 30). In August, low variability of rainfall for all regions (100%) of the study area was observed as shown in Figure 2 and this result is more or less similar to the studies by Abera et al. (2020). The highest value of the CV was found in March in the northeast part (region 2) and the lowest was recorded in August in the southeast of the sub-basin (region 8). The study's findings regarding the significant variability in annual rainfall across different regions, particularly in the eastern part of the study area exhibiting the highest variability. Region 3 shows the highest coefficient of variability in annual rainfall followed by regions 2 and 4. It also emphasizes that this variability has had adverse effects on the rain-dependent population in terms of crop production, livestock, and overall livelihood. The adverse effects on crop production, livestock, and overall livelihood emphasize the importance of understanding and addressing the impact of this variability on vulnerable communities.

The study's results suggest that, overall, the maximum and minimum temperatures in the study area exhibit lower temporal variations compared to rainfall, as illustrated in Figures 4 and 5. While most regions show lower variability in terms of minimum and maximum annual and monthly temperatures (CV < 20), specific areas with a CV exceeding 30% indicate vulnerability to natural disasters such as floods and droughts (Haile 1988; Hare 2003). The average vulnerability of approximately 58.04% throughout the year underscores the potential impact of climate-related events on the study area. Understanding the vulnerability to natural disasters and the variability in temperature patterns can be crucial for developing appropriate adaptation and mitigation strategies. The highest CV for minimum temperature was observed in January in the Megech watershed (region 1), while the lowest was recorded in December in the northeast (region 2). For maximum temperature, the highest variation was found in January in the Rib watershed (region 4), with the lowest recorded in December in the Gumara watershed (region 5). The localized vulnerability to climate-related events highlighted in the study underscores the need for region-specific adaptation and mitigation measures (Yeshitila et al. 2019; Wubneh et al. 2023). Understanding the potential impact on different regions within the study area is crucial for developing targeted strategies to address the challenges posed by climate variability.

The MK trend test and Sen's slope were applied to the study areas. The tests were applied in each month, annual mean, annual maximum, and annual minimum rainfall using a significance level of 5% and confidence level of 95% in each region (Verma et al. 2021). The obtained results are presented in Table 2. The annual rainfall in the study area was observed from 797.71 to 1,770.2 mm. The results indicate significant variations in annual rainfall across the Tana sub-basin, with the highest mean annual rainfall observed in region 6 with a value of 1,478.79 mm and the lowest mean value recorded in region 2 with a value of 985.83 mm. This heterogeneous distribution of rainfall suggests that there are significant variations in rainfall patterns within the sub-basin. Understanding these variations is crucial for assessing the potential impacts on water resources, agriculture, and ecosystems within the study area.

The MK trend test and Sen's slope were also applied to the study areas at each month, annual mean, annual maximum, and annual minimum for temperature data with a significance level of 5% and confidence level of 95% in each region. The obtained results are presented in Table 3. The annual minimum temperature in the study area ranges between 8.18 and 12.2 °C while the maximum temperature ranges between 24.00 and 29.95 °C. The results indicate significant variations in both annual minimum and maximum temperatures across the Tana sub-basin, with the highest annual maximum temperature observed in region 4 with a value of 29.95 °C and the lowest in region 6 with a value of 8.17 °C. This heterogeneous distribution of temperature further emphasizes the significant variations in minimum temperature patterns within the sub-basin. Understanding these variations is crucial for assessing potential impacts on various sectors such as agriculture, water resource management, and public health.

The analysis of Sens slope indicates that annual rainfall has increased with a non-significant trend across all regions of the study area (Weldegerima et al. 2018). Specifically, non-significant increasing trends were observed for the months of May, September, and December in all regions, while the remaining months showed non-significant increasing and decreasing trends in different regions. The magnitude of the increasing range, as indicated by Sen's slope method, ranged from 1.639 to 5.373 mm/year (Table 4 and Figure 7). Most regions showed a Sen's slope/rainfall increment value ranging from 3 to 4 mm/year, with region 5 having the highest annual increment at 5.373 mm/year, region 4 at 5.021 mm/year, and region 6 at the lowest annual rainfall increment of 1.639 mm/year (Table 4). Overall, the study found a statistically non-significant increasing trend for all parts of the sub-basin, with temporal and regional variations in rainfall increments. The findings are noted to be similar to studies conducted by Birara et al. (2020). It appears that the study indicates a general trend of non-significant increases in annual rainfall across the study area, with variations in the magnitude of the increase across different regions.

The results of the trend analysis, including the p-value from the MK test and the trend's slope (β) from Sen's estimator, are presented in Tables 5 and 6. The results of the MK test for annual minimum temperature data showed a statistically non-significant increasing trend in most regions, while regions 7, 8, and 10 showed no trend at a 5% level of significance. Similarly, a statistically non-significant increasing trend was observed for summer, spring, autumn, and winter minimum temperatures at a 5% significance level for all regions, with regions 7 and 8 in winter showing a statistically non-significant decreasing trend. The magnitude of the increasing range, as indicated by Sen's slope method, ranged from 0.001 to 0.080 °C/year (Table 5). All regions showed Sen's slope for minimum temperature increment, with region 4 having the highest annual increment and region 8 showing the lowest annual minimum temperature increment. The MK test results and Sen's slope for annual minimum temperature across the regions are presented in Table 5, indicating an increasing trend in annual minimum temperature in the Tana sub-basin.

The results of the trend analysis for annual maximum temperature data indicate a statistically non-significant increasing trend in all regions at a 5% level of significance. Similarly, a statistically non-significant increasing trend was observed for summer, spring, autumn, and winter maximum temperatures at a 5% significance level for all regions. The magnitude of the increasing range, as indicated by Sen's slope method, is from 0.043 to 0.096 °C/year (Table 6). All regions showed a Sen's slope for minimum temperature increment, with region 7 exhibiting the highest annual increment and region 4 (located in the eastern part of the study area) showing the lowest annual maximum temperature increment. The MK test results and Sen's slope for annual maximum temperature across the regions are presented in Table 6, indicating an increasing trend in annual maximum temperature in the Tana sub-basin.

Figure 7 shows the long-term trend in the minimum and maximum monthly temperature observed during the 1981–2016 periods. The temperatures ranged from 4.2 °C in December to 34.5 °C in March for the minimum and maximum temperature, respectively. These values are similar to those studied by Abera et al. (2020) for maximum temperature but the minimum temperature was recorded in February. The monthly average minimum and maximum varied from 7.66 to 11.61 °C and 28.01 to 30.56 °C, respectively.

According to the long-term average monthly temperature value, lower and higher values of monthly minimum temperature trends were observed in the Gumara watershed and Rib watershed, respectively. There was an increasing trend in Tmin and Tmax in all regions during the period 1981–2016, but it shows a non-significant trend. The increment in the magnitude of the temperature is supported in the study by Birara et al. (2020) but with a significant trend. The increment of the magnitude of temperature here is mainly related to deforestation and climate change which results from overcultivation and excessive pressure on natural resources (Anteneh 2022). The rise in temperature will lead to more evapotranspiration and is expected to increase the intensity of extreme weather events, change the amount and pattern of precipitation and impact the water resources of the study area. The variations in temperature and rainfall in the sub-basin affect agricultural activities, cropping patterns, society livelihoods and natural resources. As high population pressure and agricultural-dependent people of the study area, the rural livelihoods will be more severe since they are the direct victims of the society unless strategic measures should be taken. The vulnerability of rural households might be further aggravated if the variability of rainfall and temperature continues in the future and consequently results in drought, flood, and natural resources losses, appropriate adaptation strategies should be designed and implemented.

According to the rainfall variability and seasonality results in different regions, it is interesting to note that the highest CV in rainfall and the highest value of SI are observed in regions 3 and 4 in the eastern part of the study area. In addition, the study shows that the rainfall seasonality derived from mean PCI and the SI exhibit similar and direct graphical patterns in the study area. The regions with higher values of the SI are likely to experience rain for many months. Region 3 has the highest rainfall variability and SI, followed by region 4, while region 6 in the southern part of the study area has the lowest rainfall variability and SI. The results of the correlation between CV and SI across the regions are presented in Figure 9.

According to the PCI value of the three classes (Oliver 1980; Belay et al. 2021; Edo et al. 2021), 57.5% of rainfall data in the sub-basin were shown a strong irregular precipitation distribution (PCI > 20) and closely similar to the ones studied by Dawit et al. (2019) while 41.5% were irregular precipitation concentration. The remaining 1% of PCI values represent moderate precipitation concentration in the study area. According to the spatial and temporal distribution PCI, strong irregularity of annual, seasonal, and monthly precipitation distribution was observed. Irregularity of rainfall may lead to the occurrence of hydrological risks such as floods and drought (Yeshitila et al. 2019). In the years 1983, 1988, 1990, 2001, 2009, and 2010, hydrological risks such as flood occasions were indicated in the northeast parts of the sub-basin, and this may be the result of backflow experience from the Lake Tana to the downstream part in most of the rain season (Wubneh et al. 2023). The distribution of PCI values indicated that the distribution of rainfall in the study area was irregular spatially as well as temporally. Irregularity in the rainfall amount, intensity, onset, and offset days has a direct impact on agricultural activities, crop patterns, and livelihood. Spatially, the lowest PCI value difference (less than 2%) was observed in the year 1984 with a strong irregular rainfall occurrence. In the years 1985–1986, 1996–1997, 2014–2015, and 2019 PCI values also show small variation/range differences in all regions of the basin (Figure 10).

The study applied various statistical methods to analyze the variability and trend of climatic variables within the available rainfall and temperature data in the Tana sub-basin. The results indicated a statistically non-significant increasing trend in climatic variables across the study area. The precipitation distribution was found to be non-uniform, with a significant portion of rainfall exhibiting strong irregular distribution spatially and temporally. This irregularity in rainfall may lead to hydrological risks such as floods and drought. The study also concluded that the temporal variation of monthly rainfall is higher compared to the temporal variation of monthly temperature in the Tana sub-basin, implying that rainfall patterns are more variable and subject to change over time than temperature patterns. Given that the study area is predominantly dependent on rainfed agriculture, the variability and irregular distribution of rainfall may have significant implications for the livelihoods in the area. The study emphasizes the need for appropriate measures to mitigate and minimize the effects of climate variability on the livelihoods in the study area. The findings are expected to contribute valuable information for future similar studies.

The authors acknowledge the National Meteorology Agency (NMA) of Ethiopia for providing the observed data that were used in this study. The study was conducted on the Lake Tana sub-basin under the BRICS multilateral R&D project (BRICS2017-144) and NRF UID number 116021. The BRICS multilateral R&D project (BRICS2017-144) team is sincerely acknowledged. The authors also sincerely acknowledged Mr Sileshie Mesfin for the kind technical support and continuous follow-up that he made.

The financial assistance of the South Africa National Research Foundation (NRF) is hereby acknowledged. The study is under the grant BRICS multilateral R&D project (BRICS2017-144), the NRF UID number 116021, and the Durban University of Technology UCDG Water Research Focus Area grant.

Data cannot be made publicly available; readers should contact the corresponding author for details.

The authors declare there is no conflict.