Analysis of Trend and Periodicity in Long-Term Annual Rainfall Time Series of Nigeria - Hydrologic Time Series Analysis: Theory and Practice - page 262

Fig. 12.8. Cumulative probability distribution curve for spatial estimates of
mean annual rainfall for the 100 years (1901-2000).

zone has a total of 104 pixels, out of which only 11 or 10.6% exhibit negative
significant trends at 1% significance level. However, the negative significant
trends were observed for 18 pixels (17.3%) at 5% significance level and for
21 pixels or 20.2% at 10% significance level. Similarly, of the entire Nigeria,
3.8% exhibits negative significant trends at 1% significance level. Whereas,
negative trends were observed for 63 pixels (19.7%) at 5% significance level,
which increased to 98 pixels or 30.6% at 10% level. No pixel showed positive
significant trend at any of the three significance levels.

Table 12.4. Spatial distribution of significant negative trends at three significance
levels in the three zones and in entire study area

Significance Number of pixels with positive or negative trends

level (% within parenthesis)

Guinea Savannah Sahel Nigeria

1% 11 (10.6) 0 (0.0) 1 (1.1) 12 (3.8)

5% 18 (17.3) 28 (22.6) 17 (18.5) 63 (19.7)

10% 21 (20.2) 47 (37.9) 30 (32.6) 98 (30.6)

Figure 12.9 showed the spatial distribution of the trends at 0.5 mm/year
intervals. These values varied between -3.46 and +0.76 mm/year. About 90%
of the entire landscape exhibited negative trends while less than 10% showed
positive trends. For a better understanding of the significance of rainfall change
in Nigeria, spatial distribution of the estimated values of Mann-Kendall test-
statistics is presented in Fig. 12.10. The test-statistic values varied spatially
from -3.33 to +0.91. The spatial pattern of the changes at 10%, 5% and 1%
significance levels are vividly displayed towards the southern part of Nigeria
in the Niger Delta area and in the north-central portion of the study area. The