Sinuosity of tropical cyclone tracks in the South West Indian Ocean: Spatio-temporal patterns and relationships with fundamental storm attributes

摘要

The shapes of 268 tropical cyclone tracks in the South West Indian Ocean (SWIO) basin are investigated over the period 1977 to 2011 using an established metric for track sinuosity. Original cyclone position data from the RSMC La Reunion area of responsibility were accessed from the IBTrACS archive maintained by NOAA. Cyclone track sinuosity was measured within a GIS environment and the filtered results were normalised with a cube-root transformation function to reduce positive skew in the output sinuosity distribution. Several key findings from spatial and temporal analysis of sinuosity results may be highlighted. To assess geographical patterns, mapping tracks within designated quartile categories (straight, quasi-straight, curving and sinuous tracks) indicates which land areas in the SWIO are more likely to be affected by cyclones following sinuous tracks that are more difficult to forecast, in particular Madagascar and the islands of Reunion and Mauritius. Over the past three decades, the temporal sinuosity pattern shows a high degree of variability that will probably continue into the future. Yet crucially, a number of conspicuous episodes of relatively magnified or subdued sinuosity are recognised compared to long-term averages. This may present opportunities for identifying major climatic controls on regional anomalous cyclone migration behaviour. Within the average yearly cyclone season, the early months (September-December) have a tendency to produce more predictable straighter-moving cyclones, whereas January stands out as the singular month marking an abrupt swing towards a greater proportion (64%) of curving and sinuously-moving storms. This finding is of importance for vulnerability assessment, because a strong positive correlation is also identified between track sinuosity and cyclone longevity, such that storms steering less predictable sinuous courses are also those that tend to survive for longer durations.