3.0  ENVIRONMENTAL INFLUENCES ON ROAD BUILDING 

3.1  CLIMATE IN NAMIBIA

In order to develop cost and quality optimised road models it will be important to evaluate the environmental influences on road construction and maintenance and the knowledge about the location and properties of road building materials in Namibia.

Namibia's climate is influenced by the subtropical belt in which the whole country is situated. Namibia's precipitation is caused by a displacement of the " Southern Inter Tropical Convergence" zone in a southerly direction during the summer rainy season [12]. This ' SITC' reaches normally only the northern and central parts and is responsible for regular rains in the former and irregular precipitation in the southern regions. The aridity of most parts of Namibia is thus caused by the incomplete coverage of the 'SITC', and during the dry winter season by the general occurrence of the subtropical high pressure system. During the winter the 'SITC' zone is advancing to the equator. The influence of the Atlantic Ocean is not very far reaching into the interior and can be neglected as far as road building materials are concerned. The high moisture content in this Atlantic mist belt is, however, required for the satisfactory performance of a typical " Namibia-Adapted Technology", the " salt-gravel roads". These unique "salt roads" can only exist in a 20 km strip along the coast where the mist-caused moisture continuously interacts with the gypsum-clay pavements of these roads.

The distribution of atmospheric pressure, the precipitation and wind conditions are responsible for a parallel grouping of rainfall areas from south-west to north-east. In this general direction the precipitation as well as the reliability of occurrence of rainfalls is increasing. The change of precipitation from west to east is not as distinct as from south-west to north-east. An altitude related change of precipitation in Namibia cannot be established.

Namibia possesses three distinct climatic zones. The north-east belongs to the zone of an intermittent-moist tropical climate. This is an area of the "dry savanna" with 5 to 7,5 arid months. Rainfall can be expected between October/November to March/April. The " Mean Annual Precipitation" is concentrated in a couple of heavy showers and not distributed in regular intervals over the entire rainy season. With the exception of the border rivers Okavango and Kunene all other rivers are rivers with periodically occurring run-offs only. The mean annual precipitation ranges from approximately 450 mm to 600 mm.

The second climatic zone is occupying the largest part of Namibia. It is an area with a semi-arid dry climate with general dry rivers with very episodical run-offs. The precipitation in this semi-arid tropical climate is remarkably irregular. This region is stretching from the Kaokoveld in a south-easterly direction to the South-East-Kalahari at the Botswana border. In this area night frosts can be expected during the winter months. The general altitude of this central plateau is above 1.000 m in the average. The mean annual precipitation ranges from approximately 200 mm to 450 mm.

The third and last climatic region is situated in a south-westerly direction of the central semi-arid tropical zone. This climate is a distinct semi-desert to desert climate of the warm-moderate subtropical zone. These areas are dry during the whole year with 11 to 12 arid months. Only in exceptional good rainy seasons this region reaches a mean annual precipitation of between 100 mm and 200 mm. Under normal circumstances the ' MAP' lies at 100 mm and in the Namib and in the Orange River valley at only 50 mm.

The connection between climate and water is especially evident in arid areas like Namibia. The arid and semi-arid conditions have in the most cases not a very distinct influence on Namibia's road building materials. The most important variables affecting equilibrium moisture content in road pavement layers are material type and climate. Road construction and pavement behaviour experiences have established that moisture content increases with finer materials and with increasing Linear Shrinkage 'LS', Liquid Limit 'LL' ( Atterberg Limits) and Plasticity Index ' PI'. It has also been experienced that the equilibrium moisture content in the subgrade increases in the wetter parts of the country, whereas in the subbase and basecourse layers it appears to be independent of the climate. Under arid weathering conditions a long-term physical weathering process is responsible for a sharp weathering selection between minerals. The absence of abundant water and other biological soil forming components has the consequence that under arid weathering conditions light sediment soil layers are prevailing with many favourable road building materials. Water is important only in so far as erosion and run-off conditions of the mainly dry rivers are concerned. Most road building materials are with a few exceptions like dolerites, felspars, especially those containing biotite and muscovite, as well as mica schist and granites climate-independent.

 

3.2  FLOOD CALCULATIONS FOR NAMIBIAN CONDITIONS

 

The Namibian rainfall characteristics are associated with surprising high rainfall intensities for which provision in the design of drainage structures must be made. The run-off characteristics of most catchment areas of Namibia's river systems are generally identical with the different rainfall areas. The high evaporation in comparison with low rainfalls are responsible that all the river systems are dry rivers with irregular run-offs with the exception of the border rivers Kunene, Okavango and Orange and partly the Fish River. The Kunene, Okavango and Orange rivers are originated in areas with much higher annual precipitations than those in Namibia. Kunene and Okavango have both their origin in the central parts of the Angolan highlands and the Orange River has its origin in the mountains of Lesotho with mean annual precipitations well in excess of 1.000 mm.

The Namibian Department of Transport has over the last twenty years developed various methods for a reliable determination of design flood calculations which have proved so far to be a sound compromise between ultimate limit safety and economic factors. These design flood determination models have been developed for arid dry rivers which have, with very few exceptions, no measured run-off data. These design flood determination methods are continuously monitored against real run-off occurrences like the 50 year flood in the Khan River catchment in Usakos during 1985 and others, monitored by comparisons with test-catchment areas like the " Nubuamis Test-Catchment-Area" and by real measured run-offs at test-weirs of the " Department of Water Affairs" in some major rivers like the Swakop, Fish, Omaruru rivers and others. The most commonly used method which is used mainly for catchment areas up to 150 km2 and even larger catchment areas - to be used as an additional design control tool - will be described in some more detail. This method is represented by the " Rational Method of the Department of Transport: Namibia" [13]:

The following input data are required:

1. Area of Catchment : A (km2)
2. Length of longest Stream : L (km)
3. Average Slope of Stream : S (m/m)
4. Average Difference of Heights : H (m)
(between source of river and site of structure)
5. Roughness Factor : R
R = 0,1 for clean, compacted ground, no stones
R = 0,2 for paved areas
R = 0,3 for some grass, modest rough surface
R = 0,4 for modest grass cover
R = 0,8 for dense grass.

The Mean Annual Precipitation 'MAP' (mm) will be obtained from the newest "Mean Annual Rainfall Isohyetal Map": See table 9.

The basis for the determination of rainfall intensities is the " Time of Concentration". This is the time of a rain storm from the beginning to the peak. It is assumed that this time is equivalent to the run-off time in a river channel from the beginning of any run-off to the peak run-off. It is the time which all run-offs from the farthest points of a catchment need to contribute to the peak run-off. In the establishment of this time it must be differed between over-land-flow conditions and flow in defined river channels.

The "Time of Concentration" for overland flow is derived from the " Kerby-Formula". It is valid for overland-flow-conditions, for small, flat catchment areas as well as for the up-stream sections of river systems [14]:

Tc1 = 0,604*(R*L/S0,5)0,461 (hours)

For defined streams the "Time of Concentration" is determined by the ' HRU'-Formula:

Tc2 = (0,87*L2/1.000*S)0,385 (hours)

The calculation of the " Average Slope" is done with the following formula:

Sav = (H0,85L - H0,10L)/(1.000,00*0,75*L) (m/m)

The real "Time of Concentration" 'Tc' is determined by the weighted relations between Tc1 and Tc2.

The "Mean Annual Precipitation" (MAP) is giving for different return periods 'N' the " Maximum-Daily-Precipitation" (MDP) (mm)

The 'MDP' [15] can be calculated as follows (The graphs for 'MDP' were transformed into mathematical formulae to be used in the computer program 'RUNOFF.BAS' developed by the author):

M.D.P.= f(x) = e(-a+c*y) (mm)
y = -Log(-Log(1-1/N))
c = (6,00,5/B)*F
F = 0,47*e(-0,00435*MAP)+0,3 standard deviation
a = 0,57721*c-µ
µ = 0,93+0,39*Log(MAP)

for MAP> 86 mm : Namibian interior outside Namib

desert belt

µ = -1,26+0,98*Log(MAP)

for MAP< 86 mm : Namib desert belt

From the 'MDP' the " Estimated Point Rainfall" (mm/Tc) can be derived as follows:

Point Rainfall = 2,09*Tc*e(-a+c*y)/(1+2,88*Tc)0,92 (mm/Tc)
Intensity I = 2,09*e(-a+c*y)/(1+2,88*Tc)0,92 (mm/h)
where: c*y = (0,366*e(-0,00435*MAP)+0,234)*(-Log(-Log(1-1/N)))
a = 0,212*e(-0,00435*MAP)-0,795-0,49*Log(MAP)
for MAP> 86 mm/A
a = 0,212*e(-0,00435*MAP)+1,395-0,98*Log(MAP)
for MAP< 86 mm/A

The " Design Run-off" according to the " Rational Formula of the Department of Transport: Namibia" is as follows:

Design Run-off QD = 0,2778*C*I*A (m3/s)

where:

A = Area of Catchment (km2)
I = Point Storm Intensity (mm/h)
C = Run-off Coefficient: See table 8 [14]

 

TABLE 8  RUN-OFF COEFFICIENTS: RATIONAL FORMULA: D.o.T.

 

|==================================================================|
| COMPONENT      |         CLASSIFICATION      |    C - VALUES     |
|----------------|-----------------------------|-------------------|
|                |     Vleys and Pans (<3%)    |        0,01       |
| SURFACE/SLOPE  |    Flat Areas (3% - 10%)    |        0,06       |
|      Cs        |      Rolling (10% - 30%)    |        0,12       |
|                |        Steep (>30%)         |        0,22       |
|----------------|-----------------------------|-------------------|
|                |      Very permeable         |        0,03       |
| PERMEABILITY   |         Permeable           |        0,06       |
|      Cp        |      Semi permeable         |        0,12       |
|                |        Impermeable          |        0,21       |
|----------------|-----------------------------|-------------------|
|                |  Dense Forests,Plantations  |        0,03       |
| VEGETATION     |  Light Forests,Cultivated   |        0,07       |
|      Cv        |         Grassveld           |        0,17       |
|                |       No Vegetation         |        0,26       |
|------------------------------------------------------------------|
| REDUCTION FACTORS                                                |
|------------------------------------------------------------------|
| RETURN PERIODS (YEARS) |  100 |  50  |  25  |  20 |   15  |  10  |
|------------------------|------|------|------|------|------|------|
| REDUCTION FACTOR Rt    | 1,00 | 0,85 | 0,80 | 0,75 | 0,70 | 0,65 |
|------------------------------------------------------------------|
| REDUCTION FACTOR Rd FOR DOLOMITIC AREAS = 0,90                   |
|------------------------------------------------------------------|
| RUN-OFF COEFFICIENT C = Rt* Rd*(Cs+Cp+Cv)                        |
|==================================================================|
NOTA: D.o.T. = Namibian Department of Transport

 

TABLE 9  MEAN ANNUAL PRECIPITATION FOR 35 RAINFALL STATIONS

 

|==================================================================|
|      STATION      | MEAN ANNUAL | PERIOD | MEAN ANNUAL VALUE FOR |
|                   |  VALUE (mm) |        |  30 YEARS PERIOD (mm) |
|-------------------|-------------|--------|-----------------------|
| Rundu             |  597,8      | 1938   |                 566,6 |
| Ondangwa          |  449,0      | 1903   |                 429,8 |
| Grootfontein      |  598,9      | 1969   |                 572,7 |
| Tsumeb            |  517,7      | 1912   |                 512,9 |
| Rietfontein       |  557,4      | 1912   |                 548,6 |
| Outjo             |  397,4      | 1898   |                 417,9 |
| Otavi             |  458,0      | 1900   |                 546,2 |
| Namutoni          |  396,7      | 1903   |                 447,3 |
| Kamanjab          |  312,0      | 1931   |                 293,4 |
| Kalkfeld          |  395,7      | 1928   |                 351,4 |
| Khorixas          |  219,0      | 1956   |                 216,9 |
| Omaruru           |  276,7      | 1914*  |                 277,0 |
| Okahandja         |  345,9      | 1892   |                 369,4 |
| Usakos            |  140,5      | 1907   |                 146,8 |
| Karibib           |  227,4      | 1968   |                 216,4 |
| Gobabis           |  361,0      | 1898   |                 375,7 |
| Dordabis          |  286,3      | 1921   |                 298,9 |
| Windhoek          |  369,1      | 1890   |                 361,4 |
| Swakopmund        |   13,2      | 1900   |                  15,2 |
| Rehoboth          |  200,6      | 1884   |                 248,6 |
| Leonardville      |  237,3      | 1929   |                 249,5 |
| Aranos            |  119,7      | 1910   |                 201,5 |
| Gochas            |  108,8      | 1901   |                 185,1 |
| Mariental         |  173,6      | 1924   |                 202,8 |
| Maltahöhe         |  166,6      | 1901   |                 186,2 |
| Gibeon            |  174,5      | 1934   |                 152,9 |
| Koės              |  161,5      | 1946   |                 163,0 |
| Bethanien         |  116,2      | 1900   |                 124,6 |
| Aroab             |  145,2      | 1913   |                 171,4 |
| Keetmanshoop      |  168,5      | 1949   |                 167,0 |
| Aus               |   86,2      | 1908   |                  92,0 |
| Diaz Point        |   13,6      | 1903   |                  18,6 |
| Stampriet         |  119,6      | 1951   |                 113,2 |
| Ariamsvley        |  125,8      | 1928   |                 122,3 |
| Karasburg         |  111,2      | 1913   |                 140,6 |
|==================================================================|
NOTA: Data are valid for September 1986 and are obtained from the Meteorological Service in the Directorate: Civil Aviation: Department of Transport: Namibia. The basis is the "great book" which contains all measurements of rainfalls at official stations since 1884. All statistics have been re-calculated on the computer of the Department.

Although Omaruru's rainfall data are only recorded since 1914 it is Namibia's oldest weather station, manned by missionaries and opened in June 1882. The 30 year rainfall data are recorded for the years 1956 to 1985.

The correctness of above method for the determination of design run-offs has been proved by the following case study:

 Between 1930/31 and 1960/61 a "test-catchment-area" has been used by the Department of Water Affairs: Namibia on the farm Nubuamis north-west of Windhoek (the Nubuamis test-catchment area has Catchment Area No. 2982R02) to measure real run-off values in a small catchment area and compare them with theoretically derived run-off values like those obtained by the " Rational Method of the Department of Transport: Namibia". The Nubuamis test catchment area has got the following characteristics:

Catchment Area = 2,55 km2
Length of longest Stream = 2,20 km
Slope of Stream = 0,023 m/m
Roughness Factor = 0,3
Mean Annual Precipitation = 371 mm (between 1930/31 and 1960/61)

The highest measured peak flood occurred on 3 March 1942. The following values have been measured:

The measured peak flood was 45,5 m3/s which has been determined as a 200 years flood by the Department of Water Affairs.

The average " Mean Area Run-off" (MAR) over the whole catchment area for this flood has been 70.600 m3. This represents an average rainfall depth over the whole catchment of:

MAR/Area = 70.600/2,55 = d = 27,7 mm

The rainfall intensities for this peak storm have been measured by four rainfall gauges distributed over the test catchment area:

1) 72 mm/35 min. 2) 73 mm/45 min. 3) 58 mm/45 min. 4) 52 mm/45 min.

The average is 65 mm/45 min which is equal to 86,7 mm/h. The peak run-off occurred 55 min after the begin of the storm which represents the " Time of Concentration" Tc.

With these values the average run-off percentage which is equal to the " Run-off Coefficient" can be established:

Average Run-off Percentage = 27,7/65 = 42,6% = 0,426

All these values must now be compared with those derived by the formulae of the "Rational Method of the Department of Transport".

The "Run-Off-Coefficient" determined by table 8 gives a value of 41%. The mean value of the "Time of Concentration" between the Kerby-Formula Tc1 (0,993 h) and the HRU-Formula Tc2 (0,502 h) as calculated gives a calculated Tc of 45,2 min against the measured one of 55 min which means that the calculated Tc is on the safe side. The calculated "Intensity" according to above "Rational Method" is 78,52 mm/h against the measured one of 86,67 mm/h. With this intensity the design run-off has been calculated as: Q200 = 41,08 m3/s against the measured value of 45,5 m3/s. This represents a good correlation between the theoretically derived run-off values and measured ones under test conditions.

The "Nubuamis-Validation" of the Rational Formula of the Namibian Department of Transport has been supported by other case studies as for example the hydrological investigations for the Kalkrand Flood in 1984 when a large boxculvert on main road 38 between Kalkrand and Maltahöhe has been overtopped and those for the Usakos Flood in 1985 when a wingwall of bridge 58 over the Khan River at Usakos has been washed away as well as for the Dabib Flood in 1987 when two bridges and a number of culverts were overtopped on trunk road 1/3 between Mariental and Kalkrand and on main road 93 to Hardap with three fatalities, the first on record in the history of roads for such an occurrence. For all these case studies comparisons were made between the Rational Method of the Department of Transport and hydrological investigations of the Namibian Department of Water Affairs which compared well for all cases. Unfortunately the run-off statistics in the catchment areas of the Fish River as well as those of the Swakop River System are of considerable short duration in comparison with the Nubuamis run-off results. Other case studies to verify the new run-off determination method " Rational Method of the Department of Transport: Namibia" have been undertaken: Oanob Dam hydrology, investigating the Omuramba Omatako Dam to establish a general run-off coefficient, and the evaluation of the Uhlenhorst Storm in 1961.

It can thus be safely assumed that this method of run-off determination for arid river systems under Namibian conditions for the design of drainage structures for Namibia's roads system is the momentary best available compromise between economical and safe structures.

WB00823_.GIF (134 bytes)

[ Return to Contents ]

forward.GIF (132 bytes)