Renewable Energy Research and Development Division

Ministry of Mines, Energy, and Water Resources

P.O. Box 5285

Asmara, Eritrea

tel: 110551

Preface:

The Renewable Energy Research and Development Division of the Department of Energy is circulating this draft set of monthly rainfall maps to the branch ministries and concerned agencies in order to solicit their comments, suggestions, and corrections. In the preparation of this report, we would like to acknowledge the kind assistance and cooperation of the Department of Water Resources and the Meteorology Unit of the Civil Aviation Authority.

This rainfall analysis was performed by Robert Van Buskirk, Ph.D., Daniel Yohannes, Yonas Asmerom, and Bereket Lebasi. The latter three are research assistants recruited from the ranks of the second year Physics majors at the University of Asmara.

Introduction:

Water is a scarce and precious resource in Eritrea which is important in agriculture, environmental restoration, domestic and commercial water supply, forestry, rangeland management and biomass energy supply. National economic development and planning depends on being able to estimate the availability of water resources and their temporal and geographic distribution. Hence constant efforts need to be made to continually update and improve estimations and analysis of the rainfall resource.

To this end the Renewable Energy Research and Development Division of the Department of Energy of the Government of Eritrea is engaged in general climatic analysis in order to better understand and evaluate the solar, wind, and biomass energy resources of the country. Rainfall analysis is an important component of this effort.

Methods:

A variety of data sources were used in order to compile the monthly rainfall data set that was used. These include primarily the monthly data compiled in "Contributo alla Climatologia Dell'Altiplano Etiopico, Regione Eritrea" by Amilcare Fantoli, 1966, Ministero degli Affari Esteri, but also data sets provided by the Meteorology Section of the Civil Aviation Authority, and data were obtained from a variety of historical Italian sources obtained primarily from an archival library in Florence, Italy, and the Research and Documentation Center in Asmara.

The data was entered into spreadsheet programs, and then segregate into regions: Eastern Eritrea, Central and Northern Highlands, Southern Highlands, and Western Eritrea. For the stations in each of these regions, mean, standard deviation, skew and curtosis were calculated for the distributions of monthly rainfall amounts for each station. It was found that for each region, there was in fact a distinct relationship between the standard deviation of year to year monthly rainfall fluctuations, and the mean monthly rainfall amount (the standard deviation of rainfall fluctuations in proportional to the square root of the mean rainfall for that month), and a more complicated but apparent relationship between the skew of the distribution and the mean monthly rainfall.

A second and more important analysis of the monthly rainfall data involved calculating the years of overlap for each of the pairs of the longer-term stations in each region, and performing a students T-test for significant differences in mean rainfall for paired data. These calculations indicated which mean rainfall differences were statistically significant. And we attempted to reflect this in the isohyet maps produced. Furthermore correlation coefficients for monthly rainfall between stations was also calculated.

Furthermore, to guide extrapolation and interpolation of isohyet lines to regions were there is very little long-term data, both infrared Meteosat satellite images of cloud-top temperature, and recent reports of rainfall distributions from the Hydrometeorological Services Unit of the Water Resources Department were consulted. These sources were crucial in indicating the rainfall enhancement on the Western escarpment during June, July, and August which would otherwise not be apparent from the historic data.

The isohyet maps were produced using the average monthly rainfall amounts from 21 stations:

since these stations had reasonably long records which overlapped with other rainfall stations (typically 10-15 years or more). The area of coverage for the isohyet maps is from 36 to 40 degrees East Longitude and 14 to 17 degrees North Latitude.

CAUTIONARY NOTE:

These maps are produced with data that are biased by being mostly sampled from a relatively wet period of Eritrea's history: 1920-1955. Therefore, the monthly rainfall estimates have systematic errors in them which probably result in a 25% overestimate in mean rainfall amounts. The distribution of these maps at this time is mostly for the purpose of soliciting comments and suggestions on the maps and how they can be improved. We hope to have available in approximately six months a revised set of isohyetal maps which will include an adjustment which will account for the recent drying trend in historic rainfall totals.

Isohyet Contour Selection:

A variety of methods can be used for selecting the values of the appropriate isohyet line to show on a rainfall map. One approach is to insist on equally spaced contours. The disadvantage of this approach is that it gives high resolution in high-rainfall areas, and low resolution in low rainfall areas. We on the other hand have chosen isohyet lines using a roughly logarithmic scaling, so that both high and low rainfall regions have moderate resolution. Furthermore, we do not consider it sensible to provide discrimination between stations for whom the mean rainfall difference is not statistically significant, so we have roughly followed the proceeding convention in the number of isohyet lines:

When we say that a mean rainfall difference is 90% significant, this means that there is a 10% chance that the calculated difference between mean rainfall amounts is due to purely statistical fluctuations and the station which appears to have a lower mean rainfall actually has a higher mean rainfall.

Another implication of this style of map is that the contour spacing gives the approximate statistical error of the rainfall estimate. where the 90% confidence interval is the range that is two contour spacings wide, while the 99% confidence interval is four contour spacings wide etc. It should be kept in mind though, that other systematic errors (such as wet-year bias) will increase these errors.

Mean Rainfall Values:

The primary mean rainfall values that we used in generating the maps are shown in the proceeding table. Because of the wet-year bias in the rainfall data and the differing periods of record, several of the monthly amounts were adjusted downward slightly, and the monthly values shown in the table may not precisely match was is shown on the isohyetal (rainfall) maps. furthermore any discrepancies of 10-20% are within the statistical error of the monthly rainfall averages.

Addi Quala 0 1 7 28 40 73 256 219 69 10 8 2 713

Addi Ugri 1 3 11 32 82 59 194 191 54 10 17 3 657

Decamhare 3 4 11 30 43 47 199 166 28 2 10 4 547

Senafe 3 5 24 66 44 36 171 160 23 5 19 2 558

Zigfet 4 14 24 54 39 35 152 123 24 6 17 13 505

Abba Maitan 129 122 74 62 49 37 121 130 89 75 80 116 1084

Faghena 104 124 66 56 48 24 121 129 61 121 88 125 1067

Fil-Fil 156 187 113 82 42 16 95 81 32 77 80 134 1023

Gheleb 13 10 14 16 40 40 127 178 52 16 10 7 523

Ghinda 111 115 72 47 35 10 64 54 22 42 55 94 721

Massawa 28 29 16 13 7 1 7 8 3 15 23 38 188

Nacfa 1 7 6 22 27 25 60 92 31 12 11 1 295

Agordat 0 0 0 4 12 26 102 129 38 4 1 0 316

Barentu 5 0 0 11 23 70 148 177 77 7 3 0 521

Cullucu 0 0 2 13 18 78 205 178 86 13 6 0 599

Keren 0 0 2 14 27 52 132 138 50 4 3 0 422

Sabderat 0 0 1 4 11 32 106 151 73 12 0 0 390

Tessenei 0 0 1 5 13 38 125 147 74 11 3 0 417

Asmara 2 3 13 28 39 37 179 148 29 8 24 4 514

Belesa 3 2 5 22 32 27 133 155 30 12 8 3 405

Lat. Long. Elev. Period of Record

Addi Caieh 14° 50' 39° 20' 2461m [23-27],28-29,[30],31-36,[37-38],39-40,[41]

Addi Quala 14° 38' 38° 50' 2054m [05-06],33-40,[42-43],44-46,[47-48],49-50,

[51-53]

Addi Ugri 14° 53' 38° 49' 1972m 1895,1899-1902,[03-04],05,21,[23],24,[25,

27],28-40,[41],42-56,[57],58-63,[64-65]

Decamhare 15° 04' 39° 03' 2050m 31-34,[35],36,[37],38-40,[42],43-52,[53]

Senafe 14° 42' 39° 26' 2445m [06,23-31],32-34,[35,42],43-52,[53]

Zigfet 14° 29' 39° 29' 2595m 33-35,[36],37,[38-39],40,42-46,[47-53]

Abba Maitan 15° 42' 38° 53' 1635m 30-47,[48-50]

Faghena 15° 35' 38° 55' 1700m [26],27-36,[37],38-39,[40,50-51],52,[53]

Fil-Fil 15° 37' 38° 59' 1200m 04-07,[08],[12],13,28-36,[37-38],39-43,

[44-53]

Gheleb 15° 50' 38° 47' 1671m 27-31,[32-34],35,[36],37-38,[39,42-43],

44,[45-52]

Ghinda 15° 26' 39° 05' 960m 19,[20],[23],24-27,[28],30-34,[38],39,40,

[42-43],45,[47],48-63,[64,65],66

Massawa 15° 36' 39° 28' 5m 1885,1886-1907,12-15,[26],27,[28],29-36,

[37-41],42-66

Nacfa 16° 40' 38° 20' 1676m 05-06,[07],11-25,[26],[28-29],30-33,[34-35]

36-46,[47-49],50-61,[62],63-64,[65-66]

Agordat 15° 33' 37° 53' 633m 06,22-61,63-66

Barentu 15° 10' 37° 35' 980m [05-06,24-31],32-36,[37-38],39-40,[46-49],

[50-52]

Cullucu 14° 57' 37° 57' 870m [23-29,32,34],35-37,[38],39,[40]

Keren 15° 45' 38° 27' 1402m 19-39,[40],47,[48],49-50,[51],52-62

Sabderat 15° 27' 36° 39' 1040m 23-36,[38],39,[40]

Tessenei 15° 06' 36° 42' 615m 26-28,[29],30-40,[42],43,[44],46-47,[48],

51-56,[57],58-64,[65-66]

Asmara 15° 17' 38° 55' 2350m [1890,1893-95,1903-12],13-20,[21-22],

23-52,[53-66]

Belesa 15° 26' 38° 55' 2426m [24],25-26,37-52,[53]

In the previous table, years shown in square brackets have partially incomplete monthly rainfall records. Also, when station location is unavailable from the original source, it is taken as the location of the center of the current town locations as indicated in commonly available maps.

Conclusion:

While the attached monthly rainfall maps have significant inaccuracies, and require revision and improvement. The do represent a further step in specifying the physical geography of Eritrea. These maps are undergoing constant revision and improvement. Planned improvements include a dry-period adjustment, improved incorporation of information from short-term stations, and adjustments due to satellite rainfall estimates obtained for previously unmeasured regions (such as much of Northern Eritrea).

Again, your comments, criticisms, and suggestions are greatly appreciated.

March 2, 1996

Samuel Baire

Secretarty

Department of Energy

Ministry of Mines, Energy and Water Resources

Dear Sir:

Attached, please find a draft copy of the report Monthly Rainfall Distributions of Eritrea. Your comments and suggestions regarding the report are greatly appreciated.

Furthemore I have provided below a list of persons and institutions who might also want to commment on it. If you could send copies to these institutions, it would be quite helpful.

Sincerely,

Robert Van Buskirk, Ph.D.

Research Scientist

Renewable Energy Research and

Development Division.

cc:

Dr. Tesfai Ghirmatzion, Minister of Agriculture

Tesfai Ghebresellasi, Minister of Mines, Energy, and Water Resources

Semere Habtetsion, Head, Division of Planning and Programming, Dept. of Energ.

Haile Woldetensai, Secretary, Department of Water Resources

Hydrometeorology Unit, Department of Water Resources

Habteab Zerai, Department of Water Resources.

Head, Civil Aviation Authority

Dr. Ghebrebrhan Ogbazghi, Dean College of Science, University of Asmara

Dr. Mohamed Kheir, Dean, Faculty of Agriculture, University of Asmara

Peter McCornick, Natural Resources Consulting Engineers, P.O. Box 5260, Asmara

Land Commission

Eritrean Agency for the Environment

Azeb Tewolde, Research and Documentation Center