Ethiopia has an area of 1,112,000 square kilometers (as large as France and Spain combined), with a total population of 55 million. Of the total population, 90% of the people earn their living mainly from agricultural activity, as subsidence farmers working on 23% of the total arable land of the country.

        The Ethiopian economy is mainly based on rain-fed agriculture. Regardless of the presence of surface and groundwater resources, the failure of seasonal rains seriously affects the country's food production prospects and, in the past, has claimed millions of human and animal lives. An example of such an occurrence is the 1984 drought that resulted from the failure of the Belg (short season) rains and deficiencies in the Kiremt (main season) rainfall, aggravated by long dry spells and their untimely onset. This situation adversely affected the food intake of about 8 million people (RRC, 1985).

        However, the severity and frequency of occurrence of drought (meteorological, hydrological, and agricultural) vary for different parts of the country. Rainfall is considered as the most important climatic element that influences Ethiopian agriculture. Because of this, the rainy seasons of different parts of the country are discussed in more detail.

        The long-term mean monthly rainfall of 12 selected stations (4 from each region) is displayed in figures 2.2, 2.3, and 2.4. (The smooth curve which is overlaid over the line graphs is drawn using Harvard Graphics software.)

        The central and most of the eastern half of the country have two rainy periods and one dry period. The two rainy periods are locally known as Kiremt (June to September) and Belg (February to May), which are the long and short rainy periods, respectively. The annual rainfall distribution over this region shows two peaks corresponding to the two rainy seasons, separated by a relatively short "dry" period (Figure 2.2). The dry period, which covers the rest of the year (i.e., October to January), is known as Bega.

        The southern and the southeastern parts of Ethiopia (the regions identified by the letter C in Figure 2.1 have two distinct dry periods (December to February and September to November). The temporal distribution of rainfall over these regions shows two distinct peaks separated by a well-marked dry period (Figure 2.3).

        The western part of Ethiopia, which is identified by the letter B in Figure 2.1, has one rainfall peak during the year (Figure 2.4). The length of the rainy period decreases, and the length of the dry period increases as one goes toward the north within this region, as a result of the meridional migration of the ITCZ (Inter-Tropical Convergence Zone).

        In addition to the various atmospheric systems affecting Ethiopian rainfall, the temporal and spatial variations of rainfall discussed above could be the result of the topography and the geographical location of the country. The great East African Rift Valley (which runs northeast to southwest across Ethiopia), the mountains and highlands to the right and left of this Rift Valley, and the lowlands surrounding these mountains and highlands in every direction can be described as the country's main topographical features.

Planetary and Regional-Scale Atmospheric Systems Affecting Ethiopian Rainfall

        Ethiopia is located within the monsoon region, which is defined by surface-based observations and by consideration of atmospheric circulation patterns (Figure 3.1) (Henderson and Robinson, 1986). The main weather systems affecting Ethiopia are documented by different staff members of the National Meteorological Services Agency (NMSA).

        During Belg, the main mechanism for Ethiopia's weather patterns is the interaction between the extratropical and tropical weather systems. The Arabian ridge or anticyclone (when it is displaced to the North Arabian Sea), the penetration of deep, large-amplitude troughs in the westerlies into the lower latitudes, and the southward bend of the subtropical jet stream (STJ) at the upper levels are the major rain-producing mechanisms from February to May.

        The tropical disturbances forming over the Arabian Sea and the southwest Indian Ocean also have direct and indirect influences on Ethiopian weather during Belg and Kiremt.

        During the long rainy period (June to September), the ITCZ, the southwest Indian Ocean anticyclone (Mascarin High), the heat lows, the low-level jet (LLJ), and the South Atlantic anticyclone are the major low-level atmospheric features. The tropical easterly jet (TEJ) and the Tibetan anticyclone are the two important upper-level atmospheric features. The strength and position of these atmospheric systems vary from year to year and, so, also the rainfall activity. The ENSO episode is strongly linked with the various atmospheric systems and rainfall distribution over Ethiopia. This is briefly discussed in the next section.

The Influence of ENSO on Rain-Bearing Atmospheric Systems and Rainfall

        Various research findings have revealed the relationship between ENSO and Ethiopian rainfall and atmospheric systems. During the 1982-83 ENSO event, the TEJ was weaker and concentrated over a narrow meridional band around 60E (CSM, 1983).

        The principal cause of drought is asserted to be the fluctuation of the global atmospheric circulation, which is triggered by the SST anomalies occuring during ENSO events. These phenomena have significant impact on the displacement and weakening of the rain-producing mechanisms in Ethiopia (Haile, 1988).

        A comparative study of a drought year (1972) and a normal rainfall year (1967) over the global tropics for the northern summer months revealed (1) a weaker TEJ, (2) a weaker Tibetan high, and (3) a southeastward shift over the major circulation patterns as well as of several dynamic parameters during 1972, a major El Niño year (Krishnamurti and Kanamitsus, 1981).

       Similarly, the major rain-producing mechanism in Ethiopia and its vicinity - the ITCZ - was found to be weak, shallow, and shifted southeastward in drought years (Kruhkova, 1981; Lamb, 1978). The findings of a case study of seasonal forecasts in Ethiopia are also consistent with the above-mentioned results (Haile, 1987). These and other findings confirm that the fluctuations of atmospheric circulation, which are sometimes triggered by SST anomalies in the equatorial Pacific (ENSO) have significant impacts on the position, magnitude, and intensity of the rain-bearing systems in Ethiopia.

        The above-mentioned changes of the rain-bearing systems frequently caused meteorological, agricultural, and hydrological drought. However, the 1982-83 El Niño, which was by far the most intense one, did not produce a very dry Kiremt (Ward and Yeshanew, 1990).

        A preliminary investigation of the relationship between the equatorial eastern Pacific SSTs and Ethiopian Belg and Kiremt rains reveals that the timing and intensity of the Pacific SST anomaly must be considered while using ENSO information for forecasting purposes (Bekele, 1993).

        The following conditions are used in identifying El Niño impacts on Ethiopia:

• For an El Niño event, the positive anomaly of equatorial Pacific sea surface temperatures should last at least 12 months with a maximum value greater than or equal to 1 deg C.

• The year with the maximum observed SST anomaly is referred as an El Niño year;

• If a year is preceded or followed by an El Niño year and the positive anomaly lasted for a period of more than six months, then it is regarded as a contiguous year to the El Niño year;

• When the sea surface temperature anomaly varies similar to an El Niño year, but the maximum is somewhat less than 1 deg C (say, 0.9 deg C), it is regarded as a weak event;

        The classification of El Niño and La Niña events by their timing of occurrence was based on the timing of the significant SST increase (greater than or equal to 0.5 deg C). In the first group, the anomaly increases considerably in the period from January to June, in the second during July to December (Zang Hengfan and Wang Shaowu, 1989).

        Based on this classification and a seasonal rainfall analysis of 223 rainfall stations for the period 1969 up to 1987, the following conclusions were made about the relationship between Belt and Kiremt seasonal rainfall and eastern equatorial Pacific sea surface temperature anomalies (SSTA):

• A negative SSTA is strong associated with rainfall deficiency in Belg;

• A positive SSTA is mostly associated with normal and above-normal rainfall amounts in Belg;

• The Group One type of El Niño is always associated with a severe and widespread meteorological drought in Ethiopia during Kiremt. The 1972 and the 1987 El Niño events can be taken as examples;

• The occurrence of Group Two types of El Niño events seems to have relatively less negative effect on Kiremt rains, both in its amount and its spatial distribution. The 1982-83 El Niño event can be taken as an example;

• ENSO events may not be the only cause of meteorological drought in Ethiopia.

        Another preliminary investigation shows that the formation of intense and frequent tropical disturbances over the southeast Indian Ocean occurs simultaneously with Belg and Kiremt rainfall deficiency in Ethiopia (Bekele, 1992).

        Therefore, Ethiopian use of ENSO information, especially its seasonal forecasts, is based on existing knowledge and experience of the staff members of NMSA, as well as on the results of continuous studies and investigations. In addition to the available results of preliminary studies and research findings, ENSO information of the current year is very important for the preparation of seasonal outlooks in Ethiopia.

Sources of ENSO Information

        ENSO stands for the coupling of El Niño (oceanic component) and the Southern Oscillation (atmospheric component). El Niño originally referred to the Northern Hemisphere wintertime (December, January, February) warming of the ocean waters off the coast of Peru. This warming in some years spread to or originated in the central or eastern equatorial Pacific Ocean. The Southern Oscillation is a seesaw-like motion of surface pressure, with centers of action around the Indonesia-North Australia region and the southeastern Pacific (WMO, 1987).

        There are a number of bulletins used as sources of information for the purpose of preparing seasonal forecasts in Ethiopia. The two most widely used climate bulletins are:

• Climate Diagnostic Bulletin, produced by the Climate Prediction Center (formerly the Climate Analysis Center) of the US National Weather Service, Washington, DC;

• Weekly Climate Bulletin, also produced by the Climate Prediction Center of the US National Weather Service, Washington, DC.

• The seasonal and monthly Climate Monitoring Bulletin for the Southern Hemisphere, Bureau of Meteorology, Melbourne, Australia;

• Monthly bulletin of the Drought Monitoring Center, Nairobi, Kenya;

• Seasonal and monthly bulletin of the South African Weather Bureau;

• Rain Watch and prediction for Africa, from the African Center of Meteorological Applications for Development (ACMAD).

• Hadley Center for Climate Prediction and Research, UK.

        The seasonal forecast is mainly about rainfall and its spatial and temporal distribution. The long-range forecast unit of the NMSA prepares and issues monthly and seasonal forecasts in Ethiopia. The methods of forecasting which are applied in preparing seasonal forecasts in the NMSA are based on the following: analogue, trend analysis, statistical assessments, and teleconnections.

Analogue method

        This forecasting method is based on the assumption that a current synoptic situation will likely develop in the same way as similar past synoptic situations (WMO, 1992). Therefore, a proper selection of the analogue year is very important. ENSO information is used in this method to facilitate the selection of analogue years. After obtaining sufficient information about the status of the ENSO event of the current year, years which had the same ENSO status would be identified from past records. Then, the rainfall distribution and the synoptic features of the preseasonal months of the current year would be compared with the rainfall distribution and synoptic features of the preseasonal months of the analogue years.

        The data set used to determine analogue years includes the SSTA of the equatorial Pacific Ocean, the Southern Oscillation Index (SOI), the surface charts, daily mean sea level pressure charts (ECMWF), 500 hpa chart, 200 hpa chart, tropical cyclone frequency over the southwestern Indian Ocean, and the preseasonal and seasonal rainfall.

        Finally, an analogue year which closely resembles the current year would be chosen, based on the results of the above comparison and analysis. Then, the coming season would be anticipated to be under the categories of Above Normal, Normal, or Below Normal rainfall, based on the observed rainfall of the analogue year. Similarly, the onset and cessation of the rains will be projected.

Trend method

        Using this method, the trends of the major synoptic systems are analyzed in the preseasonal period, and the result is compared with the ideal situation.

        In some other cases, depending on the type of ENSO information available, the trend of SSTs over the central equatorial Pacific and the SOI are analyzed carefully to determine the status of an ENSO event.

        A good example to show how ENSO information could be used (with the trend method) is the 1990 Kiremt seasonal forecast. It was stated as follows: "...From [Climate Diagnostics Bulletin and Climate Monitoring Bulletin], one can see that there has been a surface warming over the central equatorial Pacific Ocean since December 1989. If the warming continues, it will likely develop into an El Niño situation. Hence, previous years which have shown the same trend were selected" (NMSA, 1990).

        In the above statement of the forecast, the information about the trend of the central equatorial Pacific SSTs is used to anticipate the trend of the major synoptic systems, assuming that the coming season will be affected by an El Niño event. This is one example of the application in Ethiopia of applying ENSO information using this method.

        On the other hand, once a forecast about the status of ENSO is at hand, it can be used in the trend analysis of the major synoptic systems to anticipate ahead of time (i.e., forecast) their intensity, magnitude, position, etc.

Statistical method

        This is an objective method of forecasting, based on a statistical examination of the past behavior of the atmosphere using regression formulas, probabilities, and other statistical measures (WMO, 1992). The use of ENSO information under this method is based on the results of previous studies by different investigators on SOI and El Niño and their effects on Ethiopia's seasonal rainfall.

Teleconnections

        Oceanic and atmospheric events at great distances from the central and eastern equatorial Pacific Ocean (i.e., the field of action), observed to occur in association with ENSO events, are considered subject to being forecast with some degree of reliability.

        The effect of ENSO events on the global atmospheric circulation in general, and on the east-west overturning of the Walker Cell in particular, alters rain-bearing synoptic systems which influence the seasonal rainfall distribution of Ethiopia. Normally the ascending limb of the Walker Cell is over Africa. During the occurrence of ENSO, the descending limb replaces it (WMO, 1985).

Achievements, Problems, and Future Prospects of NMSA in Using ENSO Information in Its Seasonal Forecasts

        As discussed earlier, the country's agricultural activity is highly dependent on seasonal rainfall performance. The deviation of the onset and cessation time of the rainy period, as well as the deviation of the seasonal rainfall amount from the long-term mean, affects agricultural output. To minimize the impacts of these events, the Ethiopian NMSA has been disseminating seasonal and monthly weather outlooks since 1987.

        The seasonal forecast in Ethiopia is in an experimental stage. So far, no seasonal forecasting model has been developed. However, investigations about the effect of ENSO on Ethiopia's seasonal rainfall and its distribution are still in progress.

        Nevertheless, the delay in transmission of sufficient ENSO information to the NMSA, the shortage of trained Ethiopian manpower specializing in long-range forecasting, and the lack of sufficient equipment are the major constraints on the development of a reliable seasonal weather or climate forecast in Ethiopia. NMSA is exerting great effort in order to solve these problems in cooperation with the WMO, UNDP, and other international organizations.

        Before discussing the success rate of seasonal forecasts of the NMSA, it would be appropriate to know about its limitations. NMSA uses four categories to compare the seasonal rainfall amount with the long-term mean (SRA stands for Seasonal Rainfall Amount, LYM for Long-Range Mean):

• Above Normal: SRA/LYM * 100 > 125%

• Normal: SRA/LYM * 100 = 125% - 75%

• Below Normal: SRA/LYM * 100 = 75% - 50%

• Much Below Normal: SRA/LYM * 100 < 50%

        Currently, the seasonal weather outlooks of NMSA are distributed to higher governmental offices where they are used for planning and programming of the national economy, to other governmental organizations like the Ministry of Agriculture, and to international organizations like USAID and UNFAO.

        ENSO is a potentially useful climate signal which may affect the atmospheric circulation in a reliably predictable way, and there are prediction schemes that show significant predictive skill. The findings of TOGA, a major ten-year research program to understand ENSO, are encouraging, and further progress is anticipated (WCRP, 1992). The future is bright. More sophisticated and effective ENSO prediction procedures are emerging rapidly. The coupled general circulation model being run by the US National Meteorological Center is one example (Cane, 1992).

        Investigations are also under way at the NMSA to improve the understanding of the detailed effects of ENSO on Ethiopia's seasonal rainfall. The combination of these two research activities will significantly upgrade Ethiopia's seasonal forecast capabilities in the future.

References

Bekele, F., 1992: The Effect of Southwest Indian Ocean Cyclones on Belg and Kiremt. NMSA mimeo. Addis Ababa, Ethiopia: NMSA.

Bekele, F., 1993: Probability of Drought Occurrence under Different Events. NMSA mimeo. Addis Ababa, Ethiopia: NMSA.

Degefu, Workineh, 1987: Some aspects of meteorological drought in Ethiopia. Drought and Hunger in Ethiopia, 23-36.

Haile, T., 1988: Causes and characteristics of drought in Ethiopia. Ethiopian Journal of Agricultural Science, 10, 85-97.

Haile, T., 1987: A case study of seasonal forecasts in Ethiopia. In: WMO RAI (Africa) Seminar on Modern Weather Forecasting (Part II), 30 November-4 December 1987. Addis Ababa, Ethiopia, 53-83.

Kruzhkova, T.S., 1981: On certain features of the atmospheric circulation in periods of drought. In: R.P. Pearce (Ed.), Tropical Droughts, Meteorological Aspects and Implication for Agriculture. Gidrometeostat: Moscow, 49-55.

Lamb, P.J., 1978: Case study of tropical Atlantic surface circulation patterns during recent sub-Saharan weather anomalies, 1976 and 1986. Monthly Weather Review, 106, 482-491.

Cane, M.A., 1993: ENSO and its prediction: How well can we forecast it? In: M.H. Glantz (Ed.), Usable Science I: Food Security, Early Warning Systems and El Niño, 43-52. Boulder, Colorado: National Center for Atmospheric Research.

Ward, M., and A. Yeshannew, 1990: Worldwide sea surface temperature and Kiremt rains in Ethiopia. NMSA mimeo. Addis Ababa, Ethiopia: NMSA.

NMSA (Long-Range Forecast Group), 1990: Seasonal Weather Outlook of Kiremt. NMSA mimeo. Addis Ababa, Ethiopia: NMSA.

RRC, 1985: The Challenge of Drought. Addis Ababa: Ethiopia: Relief and Rehabilitation Commission.

Krishnamurti, T.N., and M. Kanamitsu, 1981: Northern summer planetary-scale monsoons during drought and normal rainfall months. Monsoon Dynamics, 19-48.

WCRP, 1992: A Study of Climate Variability and Predictability. Geneva, Switzerland, WMO.

WMO, 1985: The Global Climate System: A Critical Review of the Climate System during 1982-84. Geneva, Switzerland: WMO.

WMO, 1987: The Global Climate System: 1984-86. Geneva, Switzerland: WMO.

WMO, 1992: International Meteorological Vocabulary No. 182. Geneva, Switzerland, WMO.

Zang Hengfan and Wang Shaowu, 1990: El Niño and anti-El Niño events in 1854-1987. Acta Oceanologica Sinica, 9(3), 353-362.

Main IJAS Page
Table of Contents (Vol. 2)