Rainfall Analyses

Table 20.1 shows the median rainfall for all sites according to ENSO phases as well as the correlation coefficients between rainfall and OND SST anomalies in the Niño 3 Pacific region. Out of the six sites located in the southern part of Ghana, ENSO-sea-sonal rainfall correlation was highly significant at two sites namely Axim and Akatsi and significant at Kpandu. At Axim there is as much as 500 mm difference between La Niña and El Niño events, 183.8 mm at Akatsi and 102.0 mm at Kpandu. The correlation coefficients at the other two southern sites of Accra and Akuse did not show a significant correlation but the La Niña-El Niño rainfall differences of 169.3 mm and 182.4 mm respectively were higher than the 102.0 mm observed for Kpandu where there was a significant correlation. This might be suggestive of an appreciable ENSO influence at most of the sites in the south of Ghana. ENSO influence appears weak in the middle belt of Ghana as shown by the data for Kumasi. The correlation coefficient of r = -0.32 was not significant and the observed La Niña-El Niño rainfall difference of

Table 20.1. Median seasonal rainfall (mm) and their OND SST anomaly correlation coefficients

Site

ENSO phase La Niña

Neutral

El Niño

Correlation coefficient

Axim

1 429.2

1 408.4

955.2

-0.47a

Accra

569.5

512.9

427.5

-0.29 (ns)

Ada

646.8

557.9

463.0

-0.27 (ns)

Akatsi

496.6

478.5

314.2

-0.56a

Akuse

573.3

616.0

506.0

-0.27 (ns)

Kpandu

653.4

579.2

551.4

-0.47b

Kumasi

681.0

664.0

604.0

-0.32 (ns)

Yendi

1 040.0

967.3

821.4

-0.46a

Wa

915.1

854.3

812.5

-0.26 (ns)

a Significant at 1%.

Significant at 5%. (ns) = not significant.

a Significant at 1%.

Significant at 5%. (ns) = not significant.

75 mm was smaller than that for the other sites. Out of the two northern Ghana sites, a highly significant correlation was observed for Yendi but not significant at Wa in spite of a fairly high La Niña-El Niño seasonal rainfall difference of 102.6 mm.

ENSO Effect on Simulated Peanut Yield

The results of peanut yield simulation for Akatsi are shown in Fig. 20.1. The CDFs for the planting dates were not clearly separated from each other. However, the intermediate planting date (PD2) resulted in the highest median yields for Kpedevi under all ENSO phases. At the same planting date however, median Kpedevi yields varied according to ENSO scenarios. Median yields of about 1300, 1650 and 1700 kg ha-1 were observed for El Niño, normal (or neutral) and La Niña conditions respectively (Fig. 20.1a-c). This trend is similar to that of rainfall noted earlier for Akatsi. In the case of Goronga, the CDFs for the planting dates again were not clearly separated particularly under El Niño conditions. However, PD1 resulted in the highest median yield of about 1600 kg ha-1 for the El Niño phase. PD1 again produced the maximum yield of about 2 500 kg ha-1 under La Niña conditions while the best yield of 2 600 kg ha-1 was obtained at PD2 for the neutral phase (Fig. 20.1d-f). As for Kpedevi, given the same planting date, Goronga yields varied with ENSO phase. At PD1 for example, 1600, 2 500 and 2 400 kg ha-1 yields were observed for El Niño, La Niña and neutral events respectively. El Niño median yield was at least 800 kg ha-1 less than those of La Niña and neutral phases.

Fig. 20.1. Cumulative distribution functions of simulated peanut yields at Akatsi under varying planting dates and ENSO conditions; a 7 El Niño years; b 5 La Niña years; c 12 normal years; d 7 El Niño years; e 5 La Niña years; f 12 normal years

Seasonal rainfall for El Niño phase was observed to be appreciably lower than those for La Niña and neutral phases. The La Niña-El Niño rainfall difference was 182.4 mm and that for neutral-El Niño was 164.3 mm. This vast difference in moisture availability accounted for the appreciable yield reduction during El Niño seasons. La Niña and neutral phases had high (and fairly close) seasonal rainfalls with the La Niña-neutral rainfall difference being only 18.1 mm hence the similarity in yields.

Was this article helpful?

0 0

Post a comment