4. COMPILATION OF DATA FOR THE CASE STUDY
4.5 Lake Albert long term lake levels and outflows
In order to compile a complete water level record for the period 1899 to 2008 for Lake Albert, the beginning of month gauge levels published in WMO (1982) for the years 1904 – 1978 were adjusted to end of month gauge levels in the same manner described in Section 4.4.3.
Thereafter, end of month gauge readings were extracted from the archives of the DWRM in Entebbe to complete the record up to the year 2008. However, a few intermittent gaps still remained in this record from 1982 – 1994 and in 1996. These missing levels were estimated either by linear interpolation, in cases where only a few months were missing or by a correlation relationship established between total monthly Kyoga outflows and mean month Lake Levels at Butiaba during the period 1912 – 1965, when concurrent measurements are available and of good quality. The correlation relationship is illustrated in Figure 4.10. The fitted relationship is of the form;
Butiaba gauge level (m) = 8.475 + 0.0012 * (Kyoga outflow in million cubic metres) (4.3)
Figure 4.12 Correlation between monthly Lake Albert levels at Butiaba and Lake Kyoga outflows.
Equation 4.3 (r2 = 0.90) was also used to estimate Butiaba levels for the period 1899 to 1903.
y = 0.0012x + 8.475 R² = 0.9009
8 9 10 11 12 13 14 15
0 1000 2000 3000 4000 5000 6000
Butiaba gauge (m)
Lake Kyoga outflows (MCM) Butiaba gauge and kyoga outflows
for this period are viewed as a plausible estimate. This method of estimating periods of missing Lake Albert records, using linear interpolation, and cross checking qualitatively with outflow data from upstream Lake Victoria was first suggested by Petersen et al. (2008).
4.5.2 Lake Albert outflows
The earliest record of flow measurements in the Lake Albert Basin date back from 1905 at Mongalla, which is 440 kilometres downstream of the lake. The flow at Mongalla has two components which are the contribution of direct outflow from Lake Albert and augmentation from the torrent flows. Torrent flows refers to the ungauged flow contributions in the area between Mongalla and the outlet of Lake Albert. The early records at Mongalla were taken to be representative of the outflow from Lake Albert by Hurst and Phillips (1938), who based their assessment upon a derived relationship between monthly dry season flows at Mongalla, increased by 5 % to allow for losses, and simultaneous Lake Albert levels at Butiaba. This data has been published in various volumes and supplements of The Nile Basin for the period 1904- 1957 as “Discharges of Lake Albert at Mongalla” (Hurst and Phillips, 1939; Hurst and Black, 1945; Hurst et al., 1953, 1957, 1961). Monthly outflows from Lake Albert for the period 1905 – 1978 have also been published in WMO (1982) and this record is actually a concatenation of the early records derived from flow observations at Mongalla and subsequent measurements at Panyango for the period 1948 – 1978.
Later observations of Lake Albert outflow made during the implementation of the Hydromet Project directly at the mouth from Pakwach during the period 1956 – 1974, led to the conclusion that the rating curve was unstable due to the effects of invasive aquatic vegetation and lake level influence (WMO, 1974). Measurements were therefore transferred further downstream to Panyango in 1948. The station at Panyango continued to be regularly monitored until 1978 but the record of flows is discontinuous from then till 1995 due to the prevailing insecurity in Uganda at the time. The station at Panyango was restarted in 1996 but since then gaugings have been irregular and few. Observations of outflow further downstream at Laropi commenced in 1958 but these are also discontinuous during the period 1971- 2000.
Due to political instability in Southern Sudan in 1983, flow measurements at Mongalla were suspended. To address the lack of continuity of observations at Mongallla which is a key long term monitoring station from which outflows from Lake Albert have in the past been deduced,
Petersen et al. (2008) were able to estimate missing flow data at Mongalla from the two components of Lake Albert outflow and also derived a new rating curve relating Lake Albert water levels at Butiaba to Lake Albert outflows at Mongalla, excluding estimated torrent flows.
In their study, Petersen et al. (2008), simulated torrent flows by application of the Collaborative Historical African Rainfall Model (CHARM) to a rainfall data set, for the period 1961 – 1996, in the catchment between Lake Albert and Mongalla using GIS techniques. The CHARM derived torrent flow was validated based on a calibration process where the torrent flow was assumed to be the difference between observed flow at Mongalla and observed flow at Panyango during periods of overlapping data. Their results provide a reliable update of derived Lake Albert outflows, as a separate component of observed discharges at Mongalla for the period 1979 – 1983 and calculated discharges at Mongalla for the period 1984 - 1996.
Extension of the record of Lake Albert outflows from 1996 – 2008 requires updating and application of the rating curve at Panyango based on all gaugings available from 1971 – 2009.
This rating curve is shown in Figure 4.13.
Figure 4.13 Rating curve at Panyango (1971 – 2009).
y = 380.61x - 2193.3 R² = 0.8372
800 1,000 1,200 1,400 1,600 1,800
8.40 8.80 9.20 9.60 10.00 10.40
Discharge (m3 .s-1 )
Panyango gauge (m) Panyango gaugings
A record of Lake Albert outflows for the period 1899 – 2008 is required prior to application of the Equatorial Lake Model in the methodology that has been adopted in this case study. The required record was compiled by concatenation of four data sets that were constructed as follows:
(i) Lake Albert outflows for the period 1899 – 1904 were estimated using a derived lake level outflow relationship based on gauge readings at Butiaba and directly measured discharges at Panyango (Mott MacDonald, 1998). This relationship has been compared against recently observed gaugings and was found to still be valid (Figure 4.14).
Figure 4.14 Rating used to extend Lake Albert discharge records (after, Mott MacDonald, 1998).
(ii) Adoption of the monthly outflows from Lake Albert for the period 1905 – 1978 (WMO, 1982).
(iii) Lake Albert outflows (Petersen et al., 2008), derived as a separate component of observed discharges at Mongalla for the period 1979 – 1983 and calculated discharges at Mongalla for the period 1984 - 1996.
(iv) Most recent record of Lake Albert outflows for the period 1997 – 2008 derived following application of the rating curve at Panyango illustrated in Figure 4.10.
9 10 11 12 13 14 15
250 750 1250 1750 2250
Gauge level (m)
Discharge (m3.s-1) Observed, 1948-55
Observed 1971-76 Fitted ratings Observed 1956-70 Observed 2001-2009