THE 2014-2015 CYCLONE SEASON
IN THE SOUTH-WEST INDIAN OCEAN
With a total count of eleven tropical storms, of which five developed into tropical cyclones (including two into very intense tropical cyclones), cumulating 56 days of disturbed activity (i.e. with the presence over the basin of a system at the stage of tropical storm or cyclone), the 2014-2015 cyclone season was slightly more active than average over the South-West Indian Ocean basin. But the most striking feature of the season, and which made it really specific, would have been the quite unusual typology of the tracks followed by the different named systems. One witnessed a remarkable predominance of eastward or southeastward oriented trajectories, in particular within the tropical domain. Those atypical trajectories resulted from the abnormal meteorological situation that prevailed over the basin, especially in January 2015, a month that cumulated by far the strongest disturbed activity (with no fewer than five phenomena).
That particular configuration of the trajectories was a priori rather favourable for the inhabited lands, more especially for the eastern coast of Madagascar, which as a matter of fact was exempt of any impact of a storm (for the third consecutive year). Actually, only one meteor directly hit a territory, namely tropical storm CHEDZA, which made landfall on the western coast of the Great Island. But the almost non-existence of impact did not mean that the season was completely merciful for the inhabitants of the region. Because CHEDZA alone claimed many casualties and damage, not only in Madagascar, but also in Mozambique and above all in Malawi, a country that was the most affected by the rains generated during the pre-genesis phase of the system, with induced deadly floods, the worst that the South of the country had ever undergone for 24 years.
During the 2014-2015 season, 10 tropical storms formed in the South-West Indian Ocean. In addition, one tropical cyclone (KATE) originating from the South-East Indian Ocean, in the Australian area of responsibility, inserted into the list of the endogenous phenomena, as it crossed longitude 90°East at the end of December 2014, to spend the second part of its life-time in the RSMC’s (La Réunion Tropical Cyclones Centre) area of responsibility. So that in the end 6 tropical storms and 5 cyclones evolved over the South-West Indian Ocean (a ratio between storms and mature cyclones in accordance with the normal standards).
In spite of that total amount of 11 tropical storms or cyclones observed over our basin, a number higher than the norm (which is 9), it must be considered that the activity of the season remained, for the third consecutive year, close to average, as more representative parameters of the real degree of activity of the season, such as the number of days of significant disturbed activity (number of cumulated days with the presence over the basin of a tropical storm or tropical cyclone), or the number of cyclone days (i.e. with the presence over the basin of a tropical cyclone), happened to be close to the climatological averages (slightly abovenonetheless).
In terms of activity, it can be said that the 2014-2015 season remained in keeping with the previous seasons. As a matter of fact, since the inactive 2010-2011 season, a certain stability of the disturbed activity has been noticed over the basin. Over the last three seasons in particular, the reference metrics that made it possible to characterize the degree of activity of the season thus exhibited very close values.
The mere comparison of the figures of the 2014-2015 season with those of the previous year, perfectly illustrates that high stability: the number of significant systems was rigorously identical, whether in the total number of phenomena, or in the distribution of tropical storms and cyclones (respectively 6 and 5). The mimicrywent even further, since they presented another analogy: they both counted two very intense cyclones, a common point that was far from being banal considering the fact that such a multiple occurrence is a rare event. Moreover, the cumulated number of days of significant disturbed activity remained almost unchanged (56 days instead of 55). As to the number of cyclone days, it was a little higher, rising from 18 to 21 days (in cumulated number of days – 20 in number of calendar days), which made it the second highest value over the last ten years (behind the 2006-2007 season), but without significantly standing out from the norm however. Those figures actually validated the statement that the season had an activity slightly above the norm (as the climatological averages of the cumulated number of days of disturbed activity and of the number of cyclone days are respectively 51 days – for a median set at 48 days – and 19 days).
During the 2014-2015 season, 14 low-pressure systems justified adedicated monitoring by the cyclone forecasters of RSMC La Réunion and elicited the issuance of daily bulletins, a number that also almost fitted the usual standards. There were very few cyclogeneses that aborted, since the conversion rate into mature systems was remarkable, with only two systems that failed to develop into tropical storms (N.B.: the life-cycle of system n°03 was drawing to an end when it penetrated the South-West Indian Ocean, but it had beforehand been ranked as a tropical storm, named BAKUNG, when it was still in the Indonesian area of responsibility).
The cyclone season started in mid-November, with tropical storm ADJALI, on time so to speak, since at a date corresponding to the median in terms of the occurrence dates of the first significant system (every two years, on average, the first tropical storm is named, as a matter of fact, before or after 15 November). It came to an end in mid-April, also at a date in accordance with the norm, with the withdrawing towards the southern latitudes of ex-tropical cyclone JOALANE.
During the 5 months the season had lasted, the phenomena succeeded one another fairly regularly, except for a four-weeks’ pause between the end of November and the end of December and a good three weeks lull between March and the beginning of April. But on three occasions the disturbed activity multiplied, with two phenomena coexisting simultaneously, a rather uncommon triple occurrence. Last case involved IKOLA and JOALANE, two cyclones that were present at the same time over the basin in April, something that had not been seen since April 1996. Previously, in January, a tropical storm and a very intense tropical cyclone had coexisted on two occasions, with the couples BANSI/CHEDZA, then DIAMONDRA/EUNICE successively. With those successive duos of storms both including a very intense tropical cyclone, the month of January was characterized by an intense disturbed activity, representing almost half of the total disturbed activity and even 62% of the number of cumulated cyclone days. If two very intense tropical cyclones had already developed during the previous season (with BRUCE and HELLEN), such an event remains rather infrequent. And what was definitely exceptional, was the occurrence of two very intense cyclones during the same month. Thus BANSI and EUNICE symbolised that activity peak in January. As a matter of fact one had to go back in time to more than 20 years to find a similar case. It was during the great 1993-1994 season, with cyclones LITANNE and NADIA (in March 1994).
The systems of the 2014-2015 season were distributed in a rather well-balanced way between the different areas of the basin, with yet a more significant concentration in the northeastern sector of the Marcarene Islands, due in particular to the singular and uncommon juxtaposition, almost identical, of the two cusp points and related loops followed by the trajectories of cyclones EUNICE and JOALANE.
But the most significant and unusual feature of the season was, unquestionably, the strong predominance of atypical trajectories oriented in an eastward or southeastward direction. Except for KATE and GLENDA, whose pseudo-parabolic tracks corresponded to the usual standards, practically all the other systems were associated, during their mature phases, with an eastward to southeastward motion, including at rather low latitudes, within the tropical domain (between 15 and 20°South). Probably the most symbolical archetype of those tracks that ran counter to the norm, was cyclone BANSI’s trajectory, a cyclone which never at any time headed southwestward during its life-time, an extremely rare case for that Indian Ocean part of the basin and a unique global trajectory that will go down in the history of the basin.
Those unusual eastward to southeastward tracks resulted from persistent anomalies of the general circulation which reached their culminating point in January, when the Inter Tropical Convergence Zone (ITCZ) settled at an abnormally high latitude, in connection with an exceptionally powerful monsoon flow along the African coasts, combined with strong anomalies of westerly winds present in low and mid-troposphere equatorward of the ITCZ and of the associated monsoon trough, and also with the persistence of an upper trough that had settled on a long-term basis over the southeastern part of RSMC La Réunion’s area of responsibility, between 80 and 90°East, and which worked like a “well” that sucked in the low-pressure systems towards the southeast (BANSI, DIAMONDRA and EUNICE in this case).
That quite exceptional recurrence of trajectories with unusual motion components towards the east and southeast, had several consequences. The first one was that most of the phenomena were ejected towards subtropical latitudes, with a high rate of extra-tropicalisation. The second one was that virtually all the systems that threatened inhabited lands, Madagascar and the Mascarene Islands specifically, had arrived from the west, actually a configuration that is usual for the western side of the Great Island, thus frequently impacted by the storms evolving in the Mozambique Channel, but not for the Mascarene Archipelago. As a consequence, during that season, the eastern coast of Madagascar was not under the threat of the east-westward trajectories that normally prevail on the Indian Ocean side. Hence the absence of impact of any mature system on the eastern side of the Great Island for the third consecutive year (in progress). That means an unusually enjoyable long quiet period of time for that area of the basin, which is the most exposed to cyclone hazards from the climatological point of view.
As a matter of fact, only one phenomenon directly impacted the inhabited lands, Madagascar in this case, but on the western side, namely tropical storm CHEDZA, which made landfall not far from the town of Morondava at the stage of severe tropical storm. With a rather high human toll. Because although that meteor was not particularly powerful, it claimed numerous casualties and caused great damage. Such was the case in Madagascar, where the associated rainfall occurred while the country had already undergone seasonal heavy rains thereupon generating deadly floods, but also in southern Africa, where the premises of to-be CHEDZA had beforehand brought about catastrophic floods in Mozambique, Zimbabwe and Malawi, the worst ever in the South of Malawi for 24 years (176 deaths were to be deplored).
By comparison, the Mascarene Islands were more lucky. Reunion Island was influenced by three systems, cyclone BANSI and tropical storms CHEDZA and HALIBA, but in the end without any real cyclone threat, and with consequences that were limited to heavy rains and temporary high swells. Very intense tropical cyclone BANSI was of course potentially the most dangerous, but it hopefully shifted its track eastward some 380 km to the north of La Réunion, whereas it was in a rapid intensification phase, before moving around the other islands of the archipelago at a distance. And even if Rodrigues Island felt the wind of the meteor, it opportunely escaped from the core of the cyclone.