Art. LVII.—The Histories of the Storms of the 30th January and the 16th April, 1897.

Transactions and Proceedings of the Royal Society of New Zealand, Volume 30, 1897, Page 477

Art. LVII.—The Histories of the Storms of the 30th January and the 16th April, 1897. By Major-General Schaw, R.E., C.B. [Read before the Wellington Philosophical Society, 14th July, 1897.] Plates XLII.-XLIV. During our last session I read a paper before this Society on the general subject of our New Zealand storms,* Trans. N.Z. Inst., xxix., p. 61. illustrated by a series of the weather-charts prepared at the Meteorological Office at Brisbane, under the direction of Mr. Wragge. In that paper I showed the comparatively regular sequence of cyclonic disturbances which follow one another throughout the year in these latitudes, moving from west to east, and I drew attention to some of their peculiarities. I also gave a brief statement of our present knowledge of the circulation of our atmosphere, and I ventured to bring forward some theoretical views of my own as to the origin of our cyclones and anticyclones. On this occasion I propose to bring before you in some detail the histories of the two storms which visited this North Island, and did so much damage, at the end of January and at Easter this year. But before doing so I will make a few observations on the subject generally, especially having reference to remarks made by Sir James Hector after my former paper. He observed that most of our antarctic storms appear not to be closed circuits, but great atmospheric disturbances forming incomplete circulations open towards the south. This is undoubtedly the case, and it is not easily explained. I believe, however, that the apparent anomaly is capable of explanation. The very great extent in width from west to east, often included in the circulation as we trace it southwards, and as it appears on our charts, is indeed in some degree more apparent than real, as most of our charts are on Mercator's projection, in which the meridians of longitude are shown as parallel straight lines, instead of curves converging and meeting at the pole, and hence the width of the circulation southwards is really less than is shown on the charts. Still it is often very great in our latitude, covering a thousand miles or even more, and evidence is wanting to show that a sensible completion of the circuit exists far south in the case of these wide depressions which so frequently occur. How, then, is

the circulation completed? I have little doubt that the completion of the circuit is in the upper strata of the atmosphere. The plane in which the circulation is carried on is bent upwards towards the south, or it may be regarded as not being bent down there to conform with the curved surface of the earth. The vertical component of the circulation of the air, of which I treated in my former paper, comes into play, and the circulation may be really complete, although the southern part of the circle is too high above the earth's surface to be felt as wind there. This is the probable explanation as it presents itself to my mind; but, until antarctic exploration gives us definite information by observations of cloud-motions at different heights in far southern latitudes, we can obtain no certainty on this difficult question. I will proceed now to give the history first of the great storm which was felt here on the 30th January and the following days, and which was, in fact, a combination of an antarctic storm which reached us from Hobart and of a tropical hurricane which reached us from New Caledonia. Of the sources from which I have obtained the information here compiled I will speak later on. The regular telegraphic communications failed, as the telegraphs were wrecked in the North Island by each of these storms. (See Plate XLII.) The weather-charts prepared at Brisbane by the Government Meteorologist, Mr. Wragge, show on the 26th January a cyclonic depression north-west of New Caledonia; on the 27th it had moved southwards, and the centre lay to the west of the north end of New Caledonia. The lowest barometric reading there was 28.9, but the grade was very steep, and it had developed into a tropical hurricane. On the 28th Mr. Wragge's chart indicates that the hurricane had nearly exhausted itself, but this was a mistake due to the lack of telegraphic communication with Norfolk Island (which exists with New Caledonia); the storm was in full force, but it had moved southwards. The reports furnished to Brisbane from Norfolk Island as opportunity offers, of which I have been favoured with a copy, show that at 9 a.m. on the 28th the storm had reached that island; the barometer had fallen to 29.15 with a violent east-south-east wind. The eye of the storm passed directly over the island, the wind veering from east-southeast to east, east-north-east, north-west, and thence to west-south-west. The wind dropped at 9 p.m., with barometer at 28.55, and veered from north-west with a further fall of barometer, which at 2 a.m. on the 29th stood at the extremely low level of 28.43. A violent west-south-west gale had then set in, and continued, gradually moderating, until at 9 a.m. on the 30th the weather was fine, with a south-west moderate gale and a barometric reading of 29.58. During the storm

5.45 in. of rain fell, and the sea, especially on the north side of the island, was the most terrific ever remembered there. The solid crests of the waves rose 30 ft. above high-water mark, completely destroying the pier and whaling-station. Of course, orchards, banana plantations, &c., suffered very severely. It appears that the track of the storm southwards was not perfectly straight. It approached New Caledonia from the north-west; then bent, and approached Norfolk Island from the north; bent then again and approached New Zealand from the north-west; but upon the whole the track was from north-north-west to south-south-east until it reached New Zealand. At 9 a.m. on the 29th the southerly edge of the storm had reached Cape Maria van Diemen as a moderate easterly gale, with a barometric reading of 29.58 (the same as it was at the same hour at Norfolk Island with a moderate west-south-west gale). The diameter of the disturbance was therefore about five hundred miles, or the distance between the two places. The form at that time was approximately circular or oval. As it advanced over New Zealand (North Island) it seems to have been considerably distorted, both by the irregularities of the land surface and by the opposing atmospheric pressures and currents there met with, and also to have partially filled up; for the lowest barometric readings noted in New Zealand were 28.8 at Hastings at 4 a.m. on the 30th, and 28.86 at Tauranga on the same day. At Auckland, New Plymouth, Taupo, Napier, Gisborne, and Waipiro the lowest pressure seems to have occurred on the morning of the 30th, between 4 a.m. and 9 a.m., and to have been between 28.8 and 29.15. Roughly, the centre of the storm at 9 a.m. on the 30th had assumed the form of a T, the head facing southwards from New Plymouth to Napier and Gisborne, the tail or stem stretching from Taupo to Auckland. But the track of the tropical storm was here arrested and diverted to the north-east by the presence of an antarctic cyclone which at the same time reached Cook Strait from the south and west, and by anticyclones, of which one lay to the south and east of the North Island and the other over the Tasman Sea. Of the antarctic cyclone we have full information from the logs of the warships which were crossing from Hobart about this time. It had reached Hobart on the 27th January from the west (the same day that the tropical cyclone was at New Caledonia) with a barometric reading of 29.4. On the 29th it had reached New Zealand, and, as generally happens, it had divided into two depressions, rushing northwards up the west and east coasts of the South Island, with a northerly gale on the west and a southerly gale on the east

coast. At 9 a.m. on the 30th the westerly arm of this storm had passed New Zealand and been replaced by an anticyclone of high pressure. The easterly arm, combined with the westerly, was in and south of Cook Strait, where there was a heavy gale and a terribly confused sea, caused by the meeting of the two storms, with their opposing air- and wave-motions. In what proportion the energy of this antarctic storm and the comparatively passive resistance of the anticyclones were effective in arresting the southward progress of the tropical storm and diverting it to the north-east I am unable to say, but I think that the anticyclones were largely instrumental. It is to be noticed that tropical hurricanes very rarely travel as far south as New Zealand—their progress southward seems to be limited by the band of high-pressure anticyclones which, with more or less regularity, surround our earth at about the latitude of 35° south. It has so happened, from far-reaching causes which we do not at present know, that this belt of high pressure was displaced further south than usual at the beginning of this year—at least in about our longitude—and this apparently left the course clear for the hurricane which reached us. I have not been able to trace a similar instance in past years. The circular storms which sometimes pass over the North Island, with their easterly gales and rain, are generally antarctic storms, which, owing to a displacement northwards of the barrier of anticyclones, or of a gap between them, have worked up northwards over the Tasman Sea, or, as in the last storm which caused the disastrous Hawke's Bay floods in April, the antarctic storm passed northwards through Cook Strait in an atmospheric valley between two anticylones, which lay to the east and south and to the north and west of it, and so swept across the North Island in its eastward course. This happened, unfortunately, to be its line of least resistance. The course of the antarctic storm which met the tropical storm on the 30th January was deflected to the south-southeast instead of passing on to the east over Chatham Island; so that the two storms seem to have acted upon one another as two billiard-balls would act after coming into collision in a similar manner, one passing away to the north of the anticyclone, over Chatham Island, and the other to the south of that anticyclone. A few noteworthy points connected with this storm are the following:— 1. The eye of the storm covered a very large area, embracing nearly all the central part of the North Island and a portion of the sea near Gisborne. The comparative quiet in this central part of the circulation was very beneficial to

the passengers by the “Rotomahana,” which left Napier on the evening of the 29th, before the storm had reached that latitude. On the passage to Gisborne the storm struck them as a rapidly-increasing north-east gale. When off Gisborne on the early morning of the 30th Captain Gibb thought it impossible to communicate with Gisborne, and stood out to sea in the face of what he considered the most violent gale he had ever experienced, although the waves were short, as if newly formed. At 9 a.m. the barometer had fallen to 28.9, and the gale was too severe for him to make any headway. It soon, however, began to moderate, and by noon the wind had almost ceased. The sea went down quickly, and he was able to go into Gisborne and be tendered without the slightest difficulty or inconvenience. The eye of the storm had advanced so far south that he was within its area. But soon a south-westerly gale sprung up, before which he ran until he had rounded East Cape, and before the ship reached the Hauraki Gulf the storm had passed, and there was fine weather. At the same time that the “Rotomahana” was in the calm at Gisborne, the full force of the easterly violent gale was destroying a part of the Napier Breakwater, and actually lifted a man who was sitting on the breakwater and threw him some distance, injuring him considerably. 2. I would note the effect of land in influencing the circulation. The form of the hurricane circulation corresponded markedly with the general form of the North Island, and, as we have seen, the antarctic storm was divided by the South Island. 3. Where the two storms met at Cook Strait the southeast circulation of the hurricane prevailed mainly in the Strait, the opposing north-west circulation of the antarctic storm being approximately limited by the southern shores of the Strait. 4. The hurricane travelled southwards at the rate of about five hundred miles in the twenty-four hours, or about twenty miles an hour. The antarctic storm travelled eastwards about three hundred miles on the 27th–28th, about five hundred and fifty miles across the Tasman Sea on the 28th–29th, and on the 29th–30th its motion was rather a concentration to the east of the southern island than an advance. Its progress eastward was checked by the anticyclone over Chatham Island, and its northward motion by the opposing hurricane. We may now pass on to review the history of the Easter storm. The series of charts (Plates XLIII. and XLIV.) show the storm apparently in this instance sensibly circular near the Bluff on the 14th April. We could go a little further back in its history, but it would serve no purpose. We know that it has been travelling eastward, and that

its tendency is still eastward. About its origin and its motive force we have no certain knowledge, only inferences and theories. On the evening of that day—14th April—I noticed a peculiar steely blue in the clouded sky to the south-east, which seemed to portend an approaching storm. Rain began at night, and the following day, the 15th, a southerly gale began, which increased in force, and blew all Good Friday, the 16th, with heavy rain. The rain was very exceptionally heavy in the district north and west of Napier, producing the disastrous floods from the effects of which so many are still suffering. You will notice in the weather-chart for the 15th that there was a depression—a sort of air-pond—north of Cook Strait, while to the east, the direction in which the storm was travelling, there was an extensive anticyclonic high: thus the line of least resistance for the cyclonic storm was northwards, and it rushed into the air-pond and thence forced its way across the North Island, and eventually worked round to the south-east, passing over Chatham Island, as we see in the succeeding charts. The general form of the circulation in the storm assumed on the morning of the 16th approximated to that of the storm of the 30th January. We observe the same general conformity with the shape of the North Island; but for some reason the indentation of the Bay of Plenty was greatly exaggerated, the barometric reading at Rotorua remaining high (29.8), while in the centre of the storm the average was 29.3. The lowest recorded reading was at Levin, where it was 29.2. In the centre of the storm area it was calm, while the rainfall was exceptionally heavy, with hail and lightning, but the force of the gale was very great on its circumference. Thus the schooner “Waiapu” was becalmed off Castlepoint this morning at the same time that the “Zuleika” was being driven on to the rocks in Palliser Bay by the south-east gale, and the “Pirate” was wrecked on Portland Island, at the Mahia Peninsula, by the violence of the north-east gale there. It is curious to note how the circulation from south to north was bent westwards in an elbow over the northern part of the South Island, a corresponding eastward elbow from the following anticyclone projecting across the centre of that Island. The south wind, in its westerly bend, followed the Hurunui River valley and the Otira Gorge. The outer edge of the storm included the Chatham Islands in a long shallow depression, forming a forerunner to the storm in its future course. On the 17th the shape of the circulation was not materially

altered. The depth of the central depression was slightly reduced, but extended in area to the south-east, towards which the whole storm moved a little; but it still clung mercilessly to the North Island in its western end, while the eastern end had nearly reached Chatham Islands. On the 18th the western end of the storm was at Chatham Island in its extreme depression, and only its shallow extremity still lingered over the centre of the North Island of New Zealand. I must observe that in the chart for the 16th April I have shown the isobars for 29.9 as open to the south, indicating that part of the original storm still remained down there, and that only the northern part had been cut off and formed the circular storm which we experienced; but we have really no evidence to prove this, and it is very probable that the whole of the active disturbance travelled northwards, and then to the east and south-east. In this case the isobars 29.9 should be connected by the short dotted lines near the latitude of Christchurch. One very interesting question is suggested by the behaviour of the two storms which met at Cook Strait on the 30th January. They did not unite, but apparently repelled one another. We might suppose that the opposing currents of air on the outskirts of the two storms—that of the tropical hurricane on its southern side being from the east, while that on the northern side of the antarctic storm was from the west—were the cause of this repulsion. But in the constantly-recurring cases which happen in the antarctic storms following one another we observe that, when a storm is retarded by an anticyclone to the east of it, the following storm, when it overtakes it, in some way blends with it, and reinforces it—yet the westward edge of the first storm is a south wind, while the eastward edge of the following storm which unites with it is a north wind. On the other hand, the two antarctic storms are both travelling eastward on the same circle of latitude. Whether we look on these antarctic storms as circulations open towards the south, like spokes of a great wheel with a large felloe in the region of the antarctic circle, or whether we regard them as closed circuits, we can in some measure understand how each south-and-east-going particle of air in the following storm as it comes up to the north-and-east-going current in rear of the storm it has overtaken should curve round and follow the north-and-east-going current; and that so the whole of the front edge of the following storm joins into and reinforces the storm it has overtaken, the rear edge of the following storm having of course no change of direction to perform. The track or general direction of the circulating

system is the same in both the storms, and so they unite. But in the case of the two storms of the 30th January last their tracks were nearly in opposite directions, one from the north, the other from the south, although both had an eastward tendency. I think it most probable that there is a limiting angle within which storms which collide blend and unite, but outside which they must diverge. If this be so, it was well for us that the angle of collision on the 30th January was so great; had it been within the limiting angle, and the two storms had combined, the destruction wrought would have been very much greater. Another question of great interest is, What is the condition of the atmosphere in the centre of one of these circular storms? As regards horizontal motion, it is, as we have seen, nearly still; but has it any vertical motion? Is it rising upwards; and, if so, with what degree of velocity? As I remarked in my former paper, we sorely need some accurate means for ascertaining and registering the vertical components of the motions of our atmosphere. If any inventive genius can provide such an instrument he has a certain fortune as the result of his patent. In the storm of last Easter we have an example of the way in which an antarctic storm may convert the peaceful valley between two neighbouring anticyclones into a most unpleasantly active centre of motion and destructive force. But I have observed lately two cases in which similar valleys of depression lying over the Tasman Sea between two anticyclones, on the east and west sides of that sea, have developed cyclones without any extraneous aid. We know that the friction between opposing currents of air in close proximity to one another is very small. We have noticed how closely opposing currents may approach to one another in a horizontal direction without neutralising one another, as, for instance, in Cook Strait on the 30th January last. And we often see the same fact in regard to vertical superposition of opposing currents of air by the opposing directions of motion of different strata of air. But, as we have seen must evidently be the case with regard to the opposing currents in the rear edge of a retarded antarctic cyclone and the front edge of the following cyclone which overtakes and blends with it, so there evidently are conditions under which a cyclonic circulation is set up in a valley between two anticyclones. What the favouring conditions may be which produce the phenomenon we do not know; but we may naturally suppose that one cause at least may be pressure caused by the eastward movement of the western anticyclone. My attention was drawn to the subject by the log of the “Talune” in her voyage to Sydney last month. I had a

nephew on board, who wrote to me from Sydney explaining the weather they had experienced. They left this in fine anticyclone weather with north breeze, and when they reached Sydney it was similar weather with south breeze; but during the middle passage—12th and 13th June—they passed through the southern part of a moderate cyclone with wind veering from north by east to south and rough sea. Mr. Wragge's charts give no indication of this; he had no information; but on the 16th and 17th a moderate cyclone, no doubt the same, passed round the extreme north of New Zealand, over which an anticyclone was lying. The cyclone developed in the valley was circling round the anticyclone. I had often read of such attendant cyclones circling round anticyclones, but this was the first instance that had come under my notice. Apparently its origin was as above mentioned. The other instance of a cyclone developed between two anticyclones occured about a month previously. The German warship “Olga,” on her passage from Sydney to Auckland, fell in with a similar circular storm about half-way across; but it was stationary, or moving so slowly that she passed through it into fair weather again as she neared Cape Maria van Diemen. Of this storm there were no indications in New Zealand or Australia, but the conditions were the same as in the other case—an anticyclone on either side of the Tasman Sea with a valley in the middle, and evidently in this atmospheric valley the cyclonic storm must have had its origin. As a practical question, outside the laws of storms, I was much struck, on examining the map at the time of the Hawke's Bay floods, by the very great extent of the district the drainage of which converges on Napier. It is clearly inevitable that the very fertile valleys watered by all those converging watercourses must be subject to floods; but has not the clearing of the forests from the steep and comparatively useless mountain - sides of that great fertile basin exposed the good low-lying land to much more sudden and certain disaster than if the forests had been preserved? The experience of centuries has shown the regulating influence of rough wooded country compared with cleared land in absorbing rain and delaying the flow of rain-fed watercourses. I am certain that every additional acre of bush felled on the high land surrounding that district must deteriorate in value the low-lying land to a far greater extent than anything that is gained by the clearing in extent of grazing land. I sincerely hope that the remaining bush land on the upper edges of this great basin may be carefully conserved. Whether we shall ever be able to forecast more accurately the approach of storms in these islands (than is now accomplished by Captain Edwin's admirable work) is somewhat

doubtful. The Tasman Sea lies to the west of us, and all our storms, with very rare exceptions, approach us from the west and south. We learn by telegraph the weather conditions on the other side of the sea, from Hobart to Sydney and Brisbane, but how long it will take a storm existing there to reach us, where it will strike our shores, and whether it will reach us at all, are questions shrouded in much uncertainty, depending as they do on the energy of the storm and the varying resistances it may meet with on its passage eastwards. Much, however, has been accomplished, and it is only by long-continued and patient observation, and careful comparison, that we can make further progress towards the end in view. Had Norfolk Island been in telegraphic communication with Australia or New Zealand we certainly should have been forewarned of the approach of the exceptional tropical storm of the 30th January last; but unfortunately we have no outlying island to the west of us to give us information as to the approach of our ordinary storms. Chatham Island enables us to trace the courses of storms as they leave us, and the reports from there told us that the tropical hurricane passed away far to the north of it, and the opposing antarctic storm passed away to the south of it, while the Easter storm passed right over to the island; but it could not help us in forecasts, even if in telegraphic communication. It only remains for me to return my best thanks to all who have been so good as to assist me in this investigation by the valuable observations they have communicated to me. Sir James Hector and his kind assistant Mr. Gore placed all the records furnished to the Museum at my disposal, and also the logs of H.M. warships on the passage from Hobart, which Sir James Hector had obtained through the Admiral. The Union Steamship Company authorised Captain Gibb to give me the particulars of his voyage through the storm. I am equally indebted to Mr. J. N. Williams and Mr. Galway, of Hastings; to Mr. H. Hill and Mr. D. Fox, of Napier; and to Captain Kerr and Mr. Burrows, of Tauranga; to Archdeacons Dudley and Palmer; and to Mr. Nobbs, of Norfolk Island.