Cover Image
close this book Measuring drought and drought impacts in Red Sea Province
close this folder 3. Drowght, food stress, and the flood and rainfall record for Red Sea Province. Roy Cole
View the document Summary
View the document Introduction
View the document Precipitation In Red Sea Province
View the document Flood and rainfall records: problems and possibilities
View the document A definition of drought
View the document The flood record
View the document Drought and the flood record
View the document The rainfall record
View the document Drought and the rainfall record
View the document Conclusion
View the document Discussion
View the document Limitations of the study
View the document References
View the document Appendix 3.1. Annual floods for nine khors In Red Sea Province and Khor Gash in Kassala Province.
View the document Appendix 3.2. Annual flood 2 scores for nine khors in Red Sea Province and Khor Gash in Kassala Province.
View the document Appendix 3.3. Annual rainfall in mililmetres for 19 gauging stations in eastern region, Sudan.
View the document Appendix 3.4. Annual rainfall z scores for 19 gauging stations in eastern

A definition of drought

Understanding drought in Red Sea Province is dependent on understanding land use and its variability from year to year. In Red Sea Province, with the exception of favoured places such as the Tokar and Gash Deltas, strict dichotomies between activities such as the division between agricultural and pastoral have little meaning. Gradations in involvement are such that continuous is a more appropriate descriptor than dichotomous. This relationship is a result of the variable nature of the environment. This variability has implications for any definition of drought in human terms. To define drought in human terms both agriculture and pastoralism must be considered not as absolutes but as options that change from year to year.

The figures below present the long-term coefficients of variation for each of the khors and rainfall gauging stations and are presented to illustrate the variability of floods and rainfall in the study area.


.PNG Figure 3.1. C:oefficients of variation for nine khors.


Figure 3.2. Coefficients of variation for 18 rainfall gauging stations.

The interdependencies between agriculture and pastoralism in such a variable environment are so great that it would be more proper to speak of these interdependencies as complementarities. The figure below illustrates the variable relationship between floods and principal activity in the khors of Red Sea Province. A drought, as defined below, may be considered to fall somewhere between Pastoral and Pastoral and Agricultural. Severe drought, as is defined below, falls somewhere for two years running to the right of "little". "Little" means that few resources are available for any activity but those that are available are used by livestock: there has been little annual plant production and livestock subsist on Acacia tortillis and other browse.


Figure 3.3. Land use and flood size In Red Sea Province.

Years where the flood or rainfall varies between 1 standard deviation above and below the mean are notable for their frequency and should be considered as normal variation which people expect and for which, except for the most marginal of households, they are prepared. A year of flood or rainfall greater than 1 standard deviation below the mean is equivalent to little flood or rainfall. In human terms this means the failure of one annual crop and low production of annual grasses. However, crop residues will still be produced. On Figure 3.3 above pasture is the principal land use when the flood is over 1 standard deviation below the mean and a mix of pasture and moderate agricultural production when floods are around the mean. A good crop may be harvested despite the low rainfall the closer the flood or rain is to the mean (or above it) and if the rains or floods are well spaced. The more the flood is over the mean the more agriculture is practiced. It can, as it did in 1988, become the dominant activity of the khor.

One standard deviation accounts for over two thirds of the values on either side of the mean of any distribution. Beyond one and up to two standard deviations accounts for an additional third. Beyond two and up to three standard deviations accounts for the remaining fraction. 99.7 percent of all values will be within three standard deviations of the mean. Values greater than three standard deviations from the mean are generally considered outliers and are removed from calculations of central tendency. Data collected by different measures on different scales are often compared by transforming the raw data, in this case annual floods in millions of cubic metres and annual rainfall in millimetres, into standard scores called zscores. For example, a z-score expresses each annual observation of a flood series as a deviation from the average or mean value for the entire series of values. In this form the two types of measurement are directly comparable; that is to say, rainfall data and flood data, although they are measured in completely different units can be compared on the same scale. To calculate a z-score you must first calculate the mean and the standard deviation. The standard deviation is the square root of the deviation of each value in a data series from the mean. The formula for the standard deviation is


To obtain a z-score the mean is subtracted from each value in the data series and the resulting value is divided by the standard deviation.

Defining drought in terms of standard deviations below the mean seems significant in human terms because harvest surpluses are stored from year to year (up to 6 years see Cole 1989) and because pasture produced in one year can last for up to 3 to 5 years. In the western areas of Red Sea Province where interannual variation is the greatest ERGO (1989) found that pasture was usable for at least 3 years. Thesiger (1984) reported that pasture was usable 5 years after the last rain in the Empty Quarter of Saudia Arabia Further research is needed on this topic and in modelling agricultural and pastoral production in Red Sea Province. In any case, grain storage and the use of pasture for more than one year are mechanisms (among many others) that are used to smooth environmentally-induced interannual variation in livestock and human food supplies.

The cut-off of 1 standard deviation was used for 4 khors and 17 rainfall gauging stations. The khors are: Aiterba, 'Arab, Arba'at, Baraka, and Gash. The rainfall gauging stations are: 'Agig, 'Atbara, Derudeb, Erba, Gebeit, Gebeit Mine, Haya, Kassala, Muhammed Qul, Musmar, Port Sudan, Sinkat, Suakin, Tahamyam, and Tokar.

The 1 standard deviation criterion could not be used with khors or rainfall gauging stations where the coefficient of variation was more than 100% because the lower deviation was negative. Instead of the standard deviation, any annual flood or rainfall equal to or less than 20% of the mean was considered as a drought. This was done for 4 khors and 2 rainfall gauging station. The khors are: Gwob, Kass, Tahamyam, and 'Udrus. The rainfall gauging stations are: Arba'at and Halaib.

I would like to recapitulate the definition of drought above and to extend the definition to include severe drought where animal mortality and human impacts may be said to become important.

Drought is defined as one year in which the flood is equal to or greater than 1 standard deviation below the mean, or equal to or less than 20% of the mean Hood.

Severe drought is defined as two consecutive years where floods are equal to or more than 1 standard deviations below the mean, or equal to or less than 2096 of the mean flood.