Through the visual exploration of total water surface area levels of the lakes in CAF and PERD areas, we can infer a similar pattern of variation for their water surface areas, but with no significant variation effects (Figure 3).
Upon close examination of Figure 3, changes in total water surface area can be seen in individual lakes of the CAF and PERD areas, which may indicate natural changes due to varying yearly rainfall periods.
Number Lake Region 1 Aguas Claras CAF 2 Barra CAF 3 Diquada CAF 4 Palmeiras CAF 5 Poco Redondo CAF 6 Romoalda CAF 7 T imbure CAF 8 Aceiro PERD 9 Aguas Claras PERD 10 Carioca PERD 11 Central PERD 12 Chatinha PERD 13 Bone PERD 14 Gamba PERD 15 Queiroga PERD 16 Sao Jose PERD
The PERD variable is positive and significant, indicating a positive relationship from public R&D spending to private R&D spending for Denmark, and also for Finland and Iceland, as [D.
Table 3 clearly shows that all variables are nonstationary, except for one case where the test statistic indicates a trend stationary variable (the Danish PERD variable according to both test statistics).
In each of the subpanels, two versions of the model are tested where the versions comprise the variables GDP, PERD, and GDP, INT, respectively, still including country-specific intercepts in all cases.