Climate Change Projections

Four different climate change scenarios have been used: Hadley A2 and B2 and ECHAM4 A2 and B2. Figure 9.2 shows the temperature and precipitation projections for both Hadley and ECHAM A2. The values shown to the left of the dotted line are historical data for CRU and GCM. These GCM data have not yet been normalized. To the right of the dotted line, the normalized (read: 'downscaled', see Chapter 2) GCM projections are given for the simulation periods 1961-1990, 2010-2039 and 2070-2099. The data are summarized by so-called 'whisker boxes'. The horizontal line through the boxes gives the average value. The lower and upper sides of the boxes represent the average minus and plus one standard deviation, respectively. The 'whiskers' give the minimum and maximum values found in the set, unless real outliers (points that do not fit the general distribution) were present, which are marked by diamonds (MacDonald, 1992).

The temperature projections of the Hadley A2 GCM for the historical period compare well with the historical (measured) CRU data. Also for the other projections, this good comparison holds. An upward trend of 4.5°C/100 years can be seen with an increase in variability for the long-term prediction period. The GCM overestimates precipitation by 40% over the historical period. The normalized precipitation data

Fig. 9.2. Hadley (upper graphs) and ECHAM4 (lower graphs) A2 projections for the Volta Basin, showing future simulations for temperature and precipitation against measured data.

show a small but very relevant increase of almost 10%, with a significant increase in variability over the long-term projection period. The temperature projections of the ECHAM4 A2 GCM for the historical period compare well with the historical CRU data. An upward trend of 3.6°C/100 years can be seen with an important increase in variability for the long-term projection period. The shape of this upward curve is less linear, or more exponential, than for the Hadley model, suggesting ECHAM4 predicts a larger long-term and smaller mid-term change than the Hadley model. The GCM overestimates precipitation by more than 50% over the historical period. The normalized precipitation data show no increase for the mid-term and a small increase of about 8% for the long-term. Interestingly, the model predicts a decrease in variability over the mid-term projection period.

The B2 scenarios show similar trends, but less pronounced. The temperature projections of the Hadley B2 GCM show an upward trend of 3.1°C/100 years can be seen with no major increase in variability. As one may expect, this temperature increase was less than under the A2 scenario. ECHAM4 B2 GCM shows an upward trend of 2.5°C. Both models overestimate precipitation by 40% over the historical period. The average precipitation trends for Hadley A2 and Hadley B2 are comparable.

Table 9.2. Average yearly inflow into Lake Volta (km3) together with standard deviation and coefficient of variation of the respective simulation periods for Hadley A2 and Hadley B2.

Scenario

Period

Average

sd

CoeffVar

Historical

32.8

17.1

0.52

HA2

2020-2039

41.6

14.0

0.34

HB2

2020-2039

43.8

15.4

0.35

HA2

2070-2099

37.2

19.9

0.54

HB2

2070-2099

44.0

17.6

0.40

0 0

Post a comment