Wavelet-based spatiotemporal analyses of climate and vegetation for the Athabasca river basin in Canada

Monitoring spatiotemporal changes in climate and vegetation coverage are crucial for various purposes, including water, hazard, and agricultural management. Climate has an impact on vegetation, however, studying their relationship is challenging. We implemented the Least-Squares Wavelet (LSWAVE) software for investigating trend, coherency, and time lag estimation between climate and vegetation time series. We utilized Normalized Difference Vegetation Index (NDVI) time series provided by the Terra satellite and hybrid climate time series. We found that the seasonal cycles of climate and NDVI are coherent with time delay. For the entire Athabasca River Basin (ARB), the most coherent component was the annual cycle with 84% annual coherency between vegetation and temperature and 46% between vegetation and precipitation. The annual cycles of temperature and precipitation led the ones in vegetation by about two and three weeks, respectively. Relatively lower coherency was observed in the mountainous region (upper ARB) and higher coherency in the middle ARB. From the cross-spectrograms, a clear time delay pattern was observed between the annual cycles of climate and vegetation since 2000 but not for other high-frequency seasonal cycles. The results also highlighted the advantages of LSWAVE algorithms over traditional algorithms, such as linear regression and correlation. Furthermore, we analyzed the annual land use and land cover data provided by the Terra and Aqua satellites and discussed their linkage with the climate and NDVI results.

Automated summary

Monitoring spatiotemporal changes in climate and vegetation coverage is crucial for various management purposes. This study implemented the LSWAVE software to investigate the relationship between climate and vegetation time series. The results showed that there is coherence and time delay between the seasonal cycles of climate and vegetation. The LSWAVE algorithm is advantageous for analyzing such relationships and outperforms traditional algorithms.