Mapping climate vulnerability and poverty in Africa - CGSpace Home

More documents

Recommendations

Info

We then prepared similar climate grid files, but for each combination of the five GCMs and four SRES scenarios, for the years 2015, 2030, 2045 and 2060. To do this, quadratic regressions were fitted to every pixel at the resolution of the TYN SC 2.0 data set (0.5 x 0.5° latitude and longitude) for the period 1985 to 2060 at 15-year intervals (i.e. six points). The regression fit was excellent in all areas apart from some polar regions, and the root mean square error (RMSE) was negligible. To generalize the process, we formed files of the three regression coefficients and the RMSE associated with each pixel (note that these are difference data associated with a particular combination of GCM and SRES scenario). To interpolate back to the 10-minute grids, the differences between the monthly rainfalls, temperatures and diurnal temperature ranges were calculated from the regression equations for each pixel in the coarse GCM grid. For each pixel in the 10-minute grid, the interpolated value was obtained by inverse square distance weighting; details of the algorithm can be found in Jones and Gladkov (1999) and Jones and Thornton (2000). To calculate length of growing period, we proceeded as follows. For each 10-minute pixel in Africa, the climate normals data were read from the appropriate gridded file and interpolated to daily data using the method of Jones (1986). Potential evapo-transpiration was calculated by the method of Linacre (1977). The water balance was calculated using WATBAL (see Jones (1986) for the source). This uses the method of Keig and McAlpine (1974). It calculates the available soil water, runoff, water deficiency and the actual to potential evapotranspiration ratio (Ea/Et). The source code is a simplified version of Reddy (1979). Ea/Et is calculated from a square root function that fits the three points supplied by Reddy depending on soil water holding capacity. A moderate soil water holding capacity of 100 mm was assumed for all soils. On running the water balance simulation, the number of days with Ea/Et greater than 0.5 were counted as potential growing days from day-of-year 1 to day-ofyear 365. A further restriction was placed to eliminate cold highland areas. Days with average temperature less than 9 °C were not counted as growing days even if water was not limiting. Results were output as IDRISI images for easy display and manipulation (Eastman, 2001), which involved the production of difference maps for length of growing period (LGP), rainfall and temperatures for conditions in 2020 and 2050 for each combination of GCM and SRES scenario, compared with current conditions. It should be noted that no efforts were made here to distinguish unimodal and bimodal rainfall patterns, in terms of LGP. This will 24
need to be taken into account for more detailed studies of specific areas. In addition, more refined analysis will involve estimation of water holding capacities from the FAO soils map; it would be interesting (although time consuming) to look at the differences due to the soil factor. Rainfall variability Using the same 10-minute climate normals grid for current conditions as outlined above, we used the weather generator MarkSim (Jones and Thornton, 2000) to simulate 1000 years of daily weather data for every pixel in Africa. We then calculated mean and standard deviation of annual rainfall for the pixel, and then the coefficient of variation of rainfall. Previously, we have found that estimates of CV generated using MarkSim are highly unstable, and so 1000 replicates of annual rainfall are needed. A complete simulation of Africa takes several days of continuous computer time at 10-minute resolution. There were some eight pixels for which MarkSim was not able to generate data. For these pixels, mean rainfalls were calculated from the climate normals grid, and standard deviations were estimated as the arithmetic mean of the standard deviations of the eight near neighbours of each pixel. In addition to changes in LGP to 2050, it would be very useful to be able to estimate changes in rainfall variability to the same date. This cannot be done using the methods outlined here; we can generate characteristic daily data from the climate normal grids for possible future conditions using MarkSim, but that assumes that the underlying variability of rainfall for each climate type will be the same then as it will be now. This may well not be case, and while there is some information on the direction of trends in rainfall variability to the end of this century, there is considerable uncertainty and differences exist between the various GCMs (Hulme et al., 2001). More light should be thrown on this issue in the future with the further development and application of Regional Climate Models for areas of Africa. IPCC (2001) noted that increases in mean precipitation are likely to be associated with increases in variability, and precipitation variability is likely to decrease in areas of reduced mean precipitation. Many studies have shown tendencies for inter-annual rainfall variation to increase. In many regions, including parts of Africa, interannual climatic variability is strongly related to ENSO, and thus will be affected by changes in ENSO behaviour (Hanson et al., 2006). 25
Page 1: Mapping climate vulnerability and p
Page 4 and 5: Citation Thornton PK, Jones PG, Owi
Page 6 and 7: egion of eastern Africa, the coasta
Page 8 and 9: Contents Executive summary 3 1 Back
Page 10 and 11: Tables Table 1. AOGCMs in the TYN S
Page 12 and 13: • Research to reduce uncertainty,
Page 14 and 15: 2 Project objectives and activities
Page 16 and 17: 3 Framework for the study The liter
Page 18 and 19: Figure 1. Representation of vulnera
Page 20 and 21: Conceptually, there is still a cons
Page 22 and 23: 4. Climate impacts in sub-Saharan A
Page 24 and 25: Table 1. AOGCMs in the TYN SC 2.0 d
Page 28 and 29: Mapping changes to the probability
Page 30 and 31: confidence in the ability of that G
Page 32 and 33: Figure 5. Length of growing period
Page 34 and 35: Figure 6. Coefficient of variation
Page 36 and 37: Figure 7 (B). Percentage changes in
Page 38 and 39: Figure 7 (D). Percentage changes in
Page 40 and 41: Figure 8 (B). Percentage changes in
Page 42 and 43: • For human population, we now us
Page 44 and 45: Figure 9. Farming/livelihood system
Page 46 and 47: Table 4. Country-by-system breakdow
Page 48 and 49: latter, the highland systems in sou
Page 50 and 51: Table 6. Country-by-system breakdow
Page 52 and 53: Table 8. Positive LGP changes to 20
Page 54 and 55: Figure11. Number of growing seasons
Page 56 and 57: 4.3 Uncertainties in the analysis T
Page 58 and 59: 5. Poverty and vulnerability The se
Page 60 and 61: Table 9. Vulnerability indicators u
Page 62 and 63: datasets (more information on this
Page 64 and 65: political process, civil liberties
Page 66 and 67: the major development issue facing
Page 68 and 69: scenarios. For this synthesis, the
Page 70 and 71: Table 11. Rotated component (or “
Page 72 and 73: Table 12. Synthesis of possible reg
Page 74 and 75: climate change-vulnerability hotspo
Page 76 and 77:
contain components that draw far mo
Page 78 and 79:
6.1 A survey of information needs o
Page 80 and 81:
unit; the International Red Cross;
Page 82 and 83:
predominantly in southern Africa. I
Page 84 and 85:
Table 13. Institutional preparednes
Page 86 and 87:
particularly so for government and
Page 88 and 89:
6.2 Prospects for a decision suppor
Page 90 and 91:
Figure 14. Stages of impact assessm
Page 92 and 93:
An ex ante impact assessment might
Page 94 and 95:
A second issue relates to the setti
Page 96 and 97:
A second key point that these resul
Page 98 and 99:
References AchutaRao K, Covey C, Do
Page 100 and 101:
Food and Agriculture Organisation o
Page 102 and 103:
JRL (2005). GLC 2000 (Global Land C
Page 104 and 105:
Roeckner E, Oberhuber J M, Bacher A
Page 106 and 107:
Acronyms ACTS ASARECA CGIAR DFID EN
Page 108 and 109:
Appendix 2. Candidate vulnerability
Page 110 and 111:
Appendix 3. Country by system CV of
Page 112 and 113:
Appendix 4 List of contacted instit
Page 114 and 115:
112 Note 1. Indicators of Adaptive
Page 116 and 117:
Box 1: Source: IPCC, 2001 1. Adapti
Page 118 and 119:
irrigation and quality of infrastru
Page 120 and 121:
Irrigation rate is measured by net
Page 122 and 123:
moisture, especially in southern, N
Page 124 and 125:
Bryman, A and D. Cramer . 1997. Qua
Page 126 and 127:
Note 2. South-South Cooperation (pr
Page 128 and 129:
• Effective policy frameworks in
Page 130 and 131:
vulnerable countries in 12 selected
Page 132 and 133:
130 Note 3. Climate Change and Heal
Page 134 and 135:
impacts (developing nations). These
Page 136 and 137:
most relevant for the case of HIV/A
Page 138 and 139:
emerging infectious diseases; use o
Page 140 and 141:
Within the Sub-Saharan region, the
Page 142 and 143:
geographical shifts and yield reduc
Page 144 and 145:
142 Note 4. The climate, developmen
Page 146 and 147:
Figure 2 Percentage of undernourish
Page 148 and 149:
The Africa Environment Outlook desc
Page 150 and 151:
Figure 5 Total affected in hydro-me
Page 152 and 153:
Figure 5 Present and future freshwa
Page 154 and 155:
Turner B L, Kasperson R E, Matson P
Page 156 and 157:
Note 5. The Sub-Saharan Africa Chal
Page 158 and 159:
Several useful lessons were learned
Page 160 and 161:
Note 6. The ASARECA priority settin
Page 162 and 163:
Figure 1. Development domains and a
Page 164 and 165:
Note 7. The SLP’s food-feed impac
Page 166 and 167:
Figure 2. Domains in Kenya with hig
Page 168 and 169:
Note 8. The SAKSS poverty targettin
Page 170 and 171:
168 Note 9. Simulating regional pro
Page 172 and 173:
Applications (ICASA). The results f
show all

Mapping climate vulnerability and poverty in Africa - CGSpace Home

Create successful ePaper yourself

Delete template?

Save as template?