Figure 6.5.3 Percentage of variance explained by the PET multiple linear regression for: the three spring months (left panel) and three summer months (right panel) combined
Return to the point in the document where this figure is located.
Text description of figure
The figure consists of two maps of New Zealand showing the percentage of variance explained by the PET multiple linear regression for the spring months (left) and the summer months (right). The maps show a lot of spatial detail, but there are clear regional patterns.
During the spring months, the largest explained percentage of variance (60% and over) is mainly in the east of both islands; the Gisborne region, Hawke’s Bay, Wairarapa, Marlborough down to Ashburton, and Timaru. There is a strong decrease in the variance explained going from east to west in both islands. The lowest values (below 20%) are found only in the west of both islands.
The percentage of variance explained during summer is in general less than during spring. The largest percentages are still found in the east, but are now only up to 60%, and are restricted to the Gisborne region, Wairarapa, Marlborough and some scattered regions in Canterbury. Besides the general decrease in percentage explained, the largest differences compared to spring are the low percentages (up to 40%) found in the Hawke’s Bay region. There is a small increase (10%) in percentage explained in the west of the North Island.
