Analysis of Rainfall Variability for Mekelle Meteorological Station, Northern Ethiopia (1960-2009)

Nowadays rainfall variability is threatening the world. Despite this, studies related to the determination of variations, trends and fluctuations of rainfall have little local and regional specificity. Hence, this study aimed to analyze the trend and variability of the annual and seasonal rainfall occurrence at Mekelle Airport meteorological station, Northern Ethiopia. Historical rainfall data for the period 1960-2009 of the rain gauge station were obtained from National Meteorological Service Agency. Annual, seasonal and monthly rainfall trends; onset and cessation date of kiremt rainfall; growing period length and the number of rainy days and dry spell length were considered. Standardize anomaly index (SAI) and coefficient of variance was employed to analyze the rainfall data variability using INSTAT plus, SPSS version 20 and Excel. Analysis of the historical data showed extremely severe drought happened in the area in 1984 which resulted in the death of one million people, 8 million people highly affected, and 2.3 million people food in secured. 1961 was an extremely wet year. Generally, 70%, 18% and 12% years were normal, dry and wet respectively for the last 50 years. High annual, seasonal and monthly rainfall variability was observed. Though the station showed both increasing (1984-1988) and decreasing trends (1961-1968) of annual rainfall totals a declining trend was more pronounced. Besides, one to two weeks variability on the onset and cessation date was observed. The minimum and maximum kiremt season rainy days recorded in 1987 (27days) and 1986 (86 days) respectively while the maximum dry spell days recorded in the 50 years was in 1969. Thus, devising agronomic practices that retain moisture at the plant root zone and reduce crop failure due to variability of rainfall including extended dry spell could be important.

Thus, this study was conducted to fill such a knowledge gap. Because of this, an attempt has been made to study the temporal behaviour of rainfall in Mekelle airport rain gauge station of Tigray regional state, northern Ethiopia. Therefore, the main objective of this study was to analyze the trend and variability of the annual and seasonal rainfall occurrence at Mekelle Airport station. This has been achieved by considering historical data for the period 1960-2009 of the rain gauge station.

Materials and Methods 2.1 Study area
The study was conducted in Mekelle Airport meteorological station, Northern Ethiopia. The station is located at 13 o 28'N latitude and 39 o 32' E longitude. The station is located at Mekelle the capital of Tigray regional state. The altitude of the Mekelle airport meteorological station is 2084 meters above sea level (m.a.s.l). The climate of the area is mainly characterized as a semi-arid and has flat land dominated topography. The main rainy season summer (locally called kiremt) lasts for 2 to 3 months. Moreover, the area also gets a small rain shower in the spring season (locally known as Belg) from February to May. There is great inter-annual temporal rainfall variation.

Materials and Methods
Daily rainfall data of the station were obtained from the National Meteorological Service Agency for 50 years from 1960-2009. The station was selected based on the length of a record period and the relative completeness of the data considering the maximum flexible thresholds of 10% missing values. To reconstruct the gap and to fill the missing values, data were generated following the first-order Markov chain model using INSTAT plus (v3. 6) Software.
INSTAT plus, SPSS and Excel were used to analyze and summarize the daily data into monthly, seasonal and annual totals and extreme events. The fluctuations were observed by using statistical interpretations and trends. In the variability analysis annual and seasonal rainfall patterns; standardize anomaly index (SAI); annual and seasonal rainfall trends mainly kiremt and Belg rainfall; monthly rainfall trends; onset and cessation date of kiremt rainfall; growing period length kiremt and Belg rainfall and number of rainy days and dry spell length in kiremt and Belg rainfall were considered. Civil and Environmental Research www.iiste.org ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online) Vol.11, No.9, 2019 Variability Analysis Standardized Anomaly Index (SAI) was calculated as the difference between the annual total of a particular year and the long term average rainfall records divided by the standard deviation of the long term data.

= ( − µ)
Where, Z is standardized rainfall anomaly; x is the annual rainfall total of a particular year; μ is mean annual rainfall throughout observation and Std is the standard deviation of annual rainfall for observation. This index used to examine the nature of the trends to determine the dry and wet years in the record. Tshiabukole et al. (2016), used the classification system shown in the standardize anomaly index (SAI) value produced by Mckee, et al, ( 1993) to define drought intensities resulting from the SAI. They also defined the criteria for a drought event for any of the time scales. A drought event occurs any time the SAI is continuously negative and reaches an intensity of -1.0 or less ( Table 1). The event ends when the SAI becomes positive. Each drought event, therefore, has a duration defined by its beginning and end, and an intensity for each month that the event continues. The positive sum of the SAI for all the months within a drought event can be termed the drought's magnitude. -2 and less extremely dry Precipitation Concentration Index (PCI) was also used to see the temporal pattern of rainfall (Oliver, 1980)and it was computed.
Where Pi is the rainfall amount of the ith month, PCI values below 10 indicate uniform monthly rainfall distribution; values between 11 and 20 indicate high concentrations of monthly rainfall distribution, and values of 21 and above indicate a very high concentration of monthly rainfall distribution.

Result and Discussion
Annual Rainfall Analysis Analysis of historical rainfall data using the SPI is useful for demonstrating the characteristics of a station about aridity and to monitor drought events. Therefore, SPI analysis was conducted for Mekelle meteorological station historical data. Drought is becoming a common feature of the station as part of the northern region. This historical data analysis shows in 1984 (-2.11) extremely severe drought happened in the area ( Table 2). As noted by Hayes et al, (2000) such an extremely dry condition (SPI ≤ -2) happens about two to three times in a century. This intense drought covers all regions of the country but the most affected were eastern and southern Tigray, north Wollo and North Shewa.  (Mera, 2018). Over the last 50 years, the area has also experienced severe droughts in Civil and Environmental Research www.iiste.org ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online) Vol.11, No.9, 2019 1979 (-1.55) and 2008 (-1.64) and many moderate droughts lasting several years in 2004, 1978, 2001, 2000 and 1971 along with even more droughts of shorter duration (Table 2). Whereas 1961 (3.32) and 1988 (2.05) were extremely wet years. Generally, 70% (35 years), 18% (9 years) and 12% (6 years) were normal, dry and wet respectively for the last 50 years. Agricultural, energy, environmental, transportation, and social sector were considerably impacted by the droughts.
As defined in section 2.2 PCI value 21 and above indicate a very high concentration of monthly rainfall distribution. The annual PCI value for Mekelle metrological station is 26.29 indicating the poor monthly distribution of the rainfall (Table 3). Belg rainfall in the region showed very high interannual variability (PCI > 66%). Comparing the seasonal variability of rainfall in the station, belg rainfall is more variable than the kiremt rainfall. The direct effect of high annual and seasonal variability rainfall on agricultural production could tremendously affect the livelihood of the farming community in the area.  (Table 3) shows an insignificant difference between mean and median of the annual and belg season rainfall but is asymmetrical for the kiremt season. The range in the annual, kiremt and belg season is very high. The annual and seasonal rainfall PCI value is extremely variable which is 26.29%, 28.42% and 66% for the annual, kiremt and belg seasons respectively. Based on the annual and seasonal data, one can become aware of the severe irregularities in rainfall.
In Mekelle, meteorological station rainfall is low and varies from 261mm (1984) to 1107.91 mm (1961) (Figure 2). The main rainy season (kiremt rainfall) contributes largely to the annual rainfall totals in the station, its contribution was 79 % out of the total rainfall from 1960 up 2009. Belg rainfall also makes a considerable contribution to the annual rainfall by 15%. At the station, both increasing (for example from 1984-1988) and decreasing trends (1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968) of annual rainfall totals were observed. This might be due to large inter-annual fluctuation of rainfall. For instance, at Mekelle, the 1984s were generally a dry period relative to the preceding decade and rainfall recovered to more humid conditions during the 1990s that again decreased to below long term average in 1961. Thus, the analysis that ends during the late 1971s or early 2009s might show a declining trend, whereas, if the period is extended, the trend in annual rainfall totals could reduce. On the other hand, rainfall at Mekelle showed consecutive 1 to 4 year periods with wet and dry years alternatively with no apparent trend (Figures 2, 3 and 4).   Vol.11, No.9, 2019 17 and Figure 5). The Bega is a dry season. The seasonal rainfall varied from 212.4 -912.7 mm in Kiremt, mm in the Bega ( Figure 5). From the analysis ( Figure 5) a decline rainfall trend in the Kiremt season can be observed. It was seen that the decreasing of rainfall in kiremt season could be probably because of the westerly winds lose their energy before reaching the droughtprone areas including Mekelle. The analysis also shows a high variability of rainfall within and across the seasons. The results of this study are consistent with the findings of Gedefaw et al. (2019). Kiremt rainfall variability is associated with crop production. The criterion used in this study was a rainfall of 20 mm or more accumulated over three consecutive rainy days after a specified date (in this case half June) with no dry spell greater than 7 days in the next 30 days (Bello, 2008).
Results of this study ( Figure 6) showed that the most frequent onset of kiremt rainfall was on the first of July. It was also characterized with high fluctuation of onset (June 16 -July 31); variability of the onset and retreat of rain; mainly a delay of 1 or 2 weeks in the onset; greater variability at the onset and retreat of rainfall than in the mid-season causing shortening of the rainy season and late-season drought. The variations in the onset date sometimes extended up to 70 days (10 weeks) from one year to another. Furthermore, the patterns could not be easily understood, make decisions about crop planting and related activities difficult. The end of the season was defined as the date when the available soil water content dropped to 10 mm m -1 of available water (Hadi & Tombul, 2018). This date was set based on farmers' information obtained during a preliminary survey. But in this study cessation date assigned the last date where the rain showers ended. Results of this study indicated that the median date of the end of the kiremt season was characterized by the low standard deviation (<10 days) at the station and hence, the end of the rainy season in this station is variable from 2 September iiste.org ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online) Vol.11, No.9, 2019 in 1969 to 15 September in 1990 (Figure 7). Moreover, a fluctuated trend has been observed in the cessation of kiremt rainfall in the station. Onset and cessation date have affected the growing period length. The length of the growing period was calculated as a difference between the onset date and date of the end of the season. It is the year during which rainfall distribution characteristics are suitable for crop germination, establishment, and full development. It is the year categorized as the rainy or wet season the length of which varies spatially, temporally, and with crop type. The average length of the growing period in the study region varies from 66 to 85 days. Correlation analysis of LGP with onset date and end of the rainy date showed a strong relationship in the station.  27 (1987) and 86 (1986) respectively. While the minimum and maximum Belg season rainy days were 2 (1999) and 40 (1986, 1989, 1990, and 1991) respectively indicating high variability of rainfall in the last five decades.  NMSA (2001), a dry spell is a maximum number of consecutive dry days (a day that accumulates rainfall <1 mm) were counted to determine dry spell length in kiremt season. The 50 years average dry spell during the kiremt season over the station was generally long that is 8 days and there was a variable dry spell trend (minimum 2 in 1986 to maximum 19 dry spell days in 1969) in the station. Hence, it could be helpful to devise agronomic practices that retain moisture at the plant root zone and reduce crop failure due to the extended dry spell.

Conclusions
Nowadays rainfall variability is threatening the world. Despite this, studies related to the determination of variations, trends and fluctuations of rainfall have received little attention. Thus, this study aimed to analyze the variability of rainfall occurrence at Mekelle Airport meteorological station, Northern Ethiopia. In the meteorological data history of the station extremely severe drought was recorded in 1984 which resulted in the death of one million people, 8 million people highly affected, and 2.3 million people food in secured. On the other side, 1961 was an extremely wet year. Generally, 70%, 18% and 12% years were normal, dry and wet respectively for the last 50 years. The station has high annual, seasonal and monthly rainfall variability. Though the station showed both increasing (1984)(1985)(1986)(1987)(1988) and decreasing trends (1961)(1962)(1963)(1964)(1965)(1966)(1967)(1968) of annual rainfall totals a declining trend was more pronounced. Besides, one to two weeks variability on the onset and cessation date was observed. The minimum and maximum kiremt season rainy days recorded in 1987 (27days) and 1986 (86 days) respectively while the maximum dry spell days recorded in the 50 years was in 1969. The high variability of rainfall suggests the growing importance of climate-smart agronomic practices that retain moisture at the plant root zone and reduce crop failure due to variability of rainfall including extended dry spell. Civil and Environmental Research www.iiste.org ISSN 2224-5790 (Paper) ISSN 2225-0514 (Online) Vol.11, No.9, 2019