The Climate and Weather of Australia/Chapter 12

by Henry Ambrose Hunt, Griffith Taylor, E. T. Quayle

Layout 2

 

XII.(a).—FORECASTING.

In framing official predictions of the weather the daily synoptic and isobaric chart is the chief evidence relied upon.

To forecast with any degree of accuracy requires considerable experience, a knowledge of the distribution of the normal local climatic elements, and a due appreciation of the significance of the varying features of atmospheric systems and the degree to which they are affected by latitude, longitude, the seasons of the year, and, at times, even by the prevailing conditions of the land surface over which the atmospheric systems are passing.

The methods of working in the Commonwealth Meteorological Organization are similar to those adopted in all kindred scientific institutions in other countries.

Synchronous observations are received by telegraph from a number of stations from as wide an area as possible.

The following table gives the number and equipment of stations reporting daily to the Central Bureau:—

—	9 a.m. Reports.			3 p.m. Reports.		6 p.m. Reports.
	Stations with Instruments.	Wind, Weather, and Rain Report.	Rain only (when any).	Instrument Stations.	Wind and Weather.	Rain only.
Western Australia	34	..	72	1	..	..
South Australia	18	4	72	4	..	..
Queensland	40	..	59	1	..	..
New South Wales	15	1	99	2	..	..
Victoria	45	150	10	27	5	158
Tasmania	12	..	19	1	..	..
New Zealand	3	..	..	..	..	..
New Caledonia	2	..	..	..	..	..
Norfolk Islands	1	..	..	..	..	..
Cocos Islands	1	..	..	..	..	..
Rodriguez	1	..	..	..	..	..
Batavia	1	..	..	..	..	..
Singapore	1	..	..	..	..	..
Hong Kong	1	..	..	..	..	..
	175	155	331	36	5	158

Note.—Extra 3 p.m. and 6 p.m. reports are also received at the Divisional Offices in each State.

Stations with Instruments.

Wireless.

Port Moresby (Papua)	1
Macquarie Island	1
Adelie Land	1
Ships (average)	9 to 10 weekly

The barometer readings after being reduced to sea—level, to freezing point, and for standard gravity, are plotted on a separate chart, together with wind direction and wind velocity.

All readings of the same height are then connected up by lines or isobars, thus we have invariably presented on the charts a number of contour curves which represent graphically the different pressures of the atmosphere prevailing over those portions of the earth's surface under review.

Other secondary charts are also compiled for discussion. On one is shown by figures and shading the synchronous weather throughout Australasia, the amount of cloud, rain recorded, and temperatures; on a second, the variations of pressure for the 24 hours, which indicate in a measure the probable direction of atmospheric drift; on a third, the variation in temperature for the same interval of time, and this is useful in suggesting the probable route of hot and cold waves.

All these factors are discussed together with the conditions assumed to be normal for the time of the year, and from them the forecast deductions are made.

The degree of accuracy of all forecasts issued averages 88 per cent., and it may be said that material improvement cannot be looked for until the laws which govern the apparent vagaries of movements of cyclones and anticyclones both as to rate and course, and also the, at present, disconcerting fluctuations in pressure which occasionally occur, are determined. The majority of failures in the forecasts may be attributed to want of knowledge on these two points.

Various methods have been adopted to anticipate the divergences from the usual direction of movement of anticyclones and depressions of temperate latitudes, but none have as yet proved satisfactory. It might reasonably be thought that the areas of maximum pressure change would indicate the probable path of advancing centres, or that areas of decided rise and fall in temperature might also give a clue, but the results of extended experiment have proved disappointing.

M. Gabriel Guilbert submitted a method (to the Société Météorologique de France in 1891) of anticipating the approach and progress of depressions by applying the relation of surface winds to the barometric gradient. His theories are strongly supported on parts of the Continent, and appear to have a local application, but they do not seem to apply as satisfactorily in other parts of the world where they have been tested.

Mr. Edward H. Bowie, of the Washington Bureau, has attained a high degree of accuracy in forecasting by a somewhat similar method of his own, but in his case much depends upon the personal equation of the interpreter of the system.

With regard to fluctuations in pressure, we know that such occur by actual translation of the "highs" and "lows," but it is also possible for a super-wave movement to be operating over a relatively stationary lower atmosphere without materially affecting the weather at the surface, and again a dynamical upward and downward action may be a possible explanation.

These suggestions are but tentatively submitted as a desperate attempt to account for certain marked variations of pressure from day to day without a commensurate wind or weather demonstration.

---

WINTER ANITCYCLONE.	ANTARCTIC DEPRESSION— Unfavorable to Inland Rains. No Tropical connection.	ANTARCTIC Cyclone— Passing through Bass Strait.
SUMMER ANITCYCLONE.	ANTARCTIC DEPRESSION— With Trough connection with Tropical Low Pressure. Good Inland Rains.

XII.(b).—TYPES OF WEATHER.

The following series of weather charts is given in order to illustrate isobarically and otherwise some characteristic developments in our Australian weather. In connexion with each chart or sequence of charts is given an explanatory note, necessarily somewhat brief, but probably sufficient to enable the salient features in weather progression to be grasped.

Fig. 101.

Winter Anticyclones.—This is a common type of anticyclone in winter. The centre lies near Broken Hill between latitudes 32° and 33°, and anticyclonic weather obtains over almost the whole of the continent. Hence clear skies and dry air, aided by long nights, allow of great radiation of heat from the earth's surface, and frosts are general inland south from the Tropic.

Fig. 102.

Summer Anticyclones.—8th January, 1913, provided an ordinary summer type of anticyclone covering the Bight and lying centrally over a point in about latitude 40° S.— it really passed later over or south from Southern Tasmania. The temperature differences induced by the passage of such systems along the south coast are fairly well illustrated in this case, the southerly winds in its front giving Melbourne a maximum temperature of 65°, and the north-easterly winds in its rear raising the temperature at Perth to 90°.

Fig. 103.

Antarctic Depressions unfavorable to good Inland Rains (Winter Type).—7th July, 1913. This is a typical example of an "Antarctic" with no tropical connexion. The storm has plenty of wind energy, the barometer falling rapidly southwards over Victoria and Tasmania, but the unbroken belt of high pressure lying east and west over the Continent to northward shows that in rain production it is not aided by any flow of air from the tropical belt. The rains resulting from it were practically limited to coastal areas and inland slopes facing the westerly winds.

Fig. 104.

Antarctic Depression favorable to Inland Rains (Winter Type).—18th August, 1910. In this storm the wide trough between 30.0 isobars extending northwards favoured northerly winds and large cloud development as far north as Alice Springs. Rain as far inland as William Creek was recorded next day, and over the whole drainage area of the Darling River within the next two days. Rain production was probably aided considerably by the flow of air southward in front of the trough from well within the tropical belt, the cooling due to increase of latitude favouring condensation.

Figs. 105 and 106.

Antarctic Cyclone passing through Bass Strait.—6th to 8th July, 1912.—An example of an Antarctic cyclone the centre of which passed through Bass Strait. As will be seen, the rainfall was widespread, covering practically all the country south of a line joining Streaky Bay and Brisbane, or not less than 440,000 square miles. On the Victorian highlands falls of over 1 inch were common. The slow rate of movement, about 360 miles per day, is worth noting.

Figs. 107, 108, and 109.

Cyclonic Depression Bringing General Rains to Western Australia.—8th, 9th, and 10th July, 1912.—The great cyclonic depression shown entering Western Australia on the 8th July, between Perth and Geraldton, was probably of oversea tropical origin, but whether this was so or not it not only brought a splendid general rain to all Western Australia south of the Tropic, but, maintaining a trough connexion with the tropical low—pressure belt as it moved eastward, by the 13th, it gave much rain to a long strip of inland country extending from the Territory to Northern New South Wales, as well as to South Australia and Victoria.

Figs. 110, 111, and 112.

Summer Monsoonal Rains.—6th to 18th January, 1913. Though the rainfall indicated on this chart is unusually widespread and heavy over Northern Australia, it occurs in connexion with a typical monsoonal depression, and there is no reason for regarding it as caused in any other way; that is, there is no necessity for assuming the intrusion of a tropical disturbance or any great atmospheric movement southwards. It was, however, preceded by the formation of a low pressure focus over the north-western interior of Western Australia and followed by a concentration of the rainfall over Eastern Queensland, where another slight low pressure focus was formed. The latter is shown by the chart of the 8th.

CYCLONIC DEPRESSION ENTERING BETWEEN CAPE LEEUWIN AND SHARK BAY
BRINGING GENERAL RAINS TO WESTERN AUSTRALIA.

SUMMER MONSOONAL RAINS-NORTHERN AUSTRALIA.

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

ANTARCTIC CYCLONE— Passing over Inland Victoria, with general rains.	INLAND WINTER RAINS— Of Tropical origin and Cyclonic ending.	TROPICAL RAIN STORM— Moving from the Territory Southward into South Australia.

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

Figs. 113 AND 114.

Antarctic Cyclone passing inland.—1st and 2nd September, 1911. The heavy and extensive inland rain which accompanied this inland cyclone was the only good spring rain experienced in 1911, and was the salvation of the crops over South Australia, Victoria, and much of New South Wales. The chart of 1st September shows it as an ordinary Antarctic disturbance, but with an exceptionally good trough extending northwards in front of which there was a wide and unbroken flow of northerly winds from the Gulf of Carpentaria into South Australia and Victoria. It is possible that the condensation induced by the transfer of such a large body of air into higher latitudes was mainly responsible for the cyclonic development and heavy rains.

Figs. 115 AND 116.

Inland Winter Rains of Tropical Origin and Cyclonic ending.—During June and July, 1912, no less than four rain storms originated in the northern interior of the continent, two of which developed enormously and gave heavy rains to the whole eastern half of the continent. The charts of 22nd and 24th June show stages in the barometric development of the second of these storms, warning of which was first given on the 21st by the occurrence of thunderstorms in the far north-west of Queensland. Next day these were occurring over a much greater area, extending northwards beyond Powell's Creek and southwards to Broken Hill, the isobars at the same time dipping southwards towards the Bight, where previously barometers had read up to 30.5 inches. During this and the next day its rain production was at its maximum. By the 24th, a very symmetrical and fairly intense cyclone was centred near Adelaide. This afterwards rapidly lost energy both as regards pressure and rain production, and moved slowly eastwards over Victoria.

It is worthy of note that in this case, as in that of many other storms of tropical origin, the cloud and rain development were always in advance of the isobaric showing clearly the convectional origin of the cyclone which followed.

Figs. 117 AND 118.

Tropical Rain Storm moving Southward from the Territory into South Australia.—Examples of this type of storm are not very numerous, but the fact that they occur almost as readily in the winter as in the summer months shows that they are not in any sense monsoonal. Perfect examples of this occurred on the 31st July, 1908, and 6th-8th June, 1907. They all gave splendid rains over the central parts of the continent. That of 26th February—5th March, 1910, was, however, the most remarkable in that and in every other respect.

On the 25th February heavy rain was general as far south as Powell's Creek, and this was accompanied by the formation of a definite low-pressure centre south from Port Darwin. This slowly moved due southwards, accompanied by a tremendous rainfall amounting at many inland stations to very nearly the annual average. By the 5th March, South Australia was deluged, and the storm, which then became very definitely cyclonic, moved off over Victoria towards the S.E., giving rainfalls in low-lying plain country of 4 or 5 inches, and up to 12 inches amongst the mountains.

Figs. 119 and 120.

Willy-willy affecting Coastal Parts only.—The tropical cyclones, which during the summer half of the year are occasional visitants to the northwest coast of West Australia and known there as "willy-willies," generally follow paths tending to carry them inland. But in most cases their identity is then speedily lost, barometric intensity and wind energy disappearing and the storm becoming merged into an ordinary "monsoonal dip." In a few cases they keep out to sea, and, moving southwards, may even round Cape Leeuwin. In the case of the very violent one shown centred near Onslow on the 7th February, 1911, apart from the heavy local coastal rains of the 6th and 7th, the rest of the State did not benefit, and the storm either moved off seawards or broke up altogether.

Figs. 121 and 122.

Willy-willy moving Inland.—On the 19th January, 1909, is another willy-willy not quite so intense as that of the 7th February, 1911, but occurring with almost the same general pressure distribution. Yet this one moved inland, and by the 21st January had reached the Murchison gold-fields, having lost but little of its intensity, and bringing very heavy rain to the Murchison and Northern Coolgardie gold-fields. It then moved slowly eastward across the continent with diminished intensity, but still accompanied by rain even to the Queensland coast.

Figs. 123 and 124.

Gales through Bass Strait.—Westerly gales naturally occur whenever the barometric fall southwards is rapid and the isobars lie approximately east and west through the strait. The most frequent cause of this is the passage of an Antarctic low, not only intense but on a large scale over Southern Tasmania or the ocean waters south of it. Such is the storm shown on the charts of 13th and 14th September, 1912. The chart of the 13th shows what may be described as a "square-headed" low, the squareness being due to the formation of a second trough closely following the first. Such a development in an energetic disturbance invariably means severe weather, especially along the coast line, with hail showers and not infrequently thunder. The weather of the 13th was no exception. The first eighteen days of this month were exceptionally boisterous, owing to the prevalence of this type of storm.

WILLY WILLY RETURNING OCEANWARDS.	WILLY WILLY MOVING INLAND.	WESTERLY GALES THROUGH BASS STRAIT.

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

EASTERLY GALES THROUGH BASS STRAIT.	ANTICYCLONE FORCED SOUTHWARD OVER TASMANIA BY CYCLONE OFF NEW SOUTH WALES COAST.	COLD SNAP, SOUTH-EAST AUSTRALIA.

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

Figs. 125 and 126.

Easterly Gales through Bass Strait.—These also occur occasionally though only very rarely, and, of course, when the barometric gradient is reversed, the most common case being an intense monsoonal cyclone or dip over New South Wales opposed by a strong anticyclone over Tasmania. 12th and 13th January, 1911, provided a case of this kind, the weather in Bass Strait and even in Port Phillip being exceptionally severe. Very rough weather also prevailed on the New South Wales coast.

Figs. 127 and 128.

Anticyclone forced Southwards over Tasmania by Cyclone off New South Wales Coast.—The warm waters off the east coast of Australia are very favorable to vigorous cyclonic development, and many depressions reaching these waters after a journey across the continent, during which their barometric effect was very feeble, give rise to great displays of energy. But whatever their origin, whether it be directly from the tropical seas to northward, or monsoonal depressions inland, or even from the heads of Antarctic disturbances, they show a marked tendency to cling for days at a time to the coastal waters. This appears to have a twofold effect upon the following anticyclone—(a) it is partly forced southwards; (b) it is merged into a semi-circular high-pressure ridge built up on its southern sides by the cyclone and extending from southern New Zealand across the ocean to Tasmania, and then north-west or north towards the Northern Territory or the Gulf of Carpentaria. The cyclone shown on the chart of 15th July, 1912, apparently originated near Alice Springs on the 12th, where thunderstorm rains were then falling and evidence of the beginning of a separate low centre being given, became cyclonic off Port Macquarie on the 14th, and held the "High" in the position shown for four days (14th to 17th inclusive).

Figs. 129 and 130.

Cold Snap (South-eastern Australia).—The cold weather of 9th and 10th October, 1910, was phenomenal. It began almost simultaneously in South Australia and Victoria on the 9th, when heavy snowfalls during the afternoon were almost general from Spencer Gulf to Gabo Island.

The chart of the 10th, with its cyclonic centre off Gabo Island and steep gradients, indicates very unsettled and rough weather conditions, which were sufficiently illustrated in Melbourne during the early morning by heavy hail storms, vivid lightning, and sharp peals of thunder. Moreover, the great length of north and south isobars which overlay Eastern Australia and extended through at least 30° latitude favoured great transference of cold air from Antarctic seas, even into tropical Queensland. It is doubtful, however, if it can be taken as sufficient demonstration of the cause of the heavy snowfalls and exceptional cold, the snowfall and greatest drop in temperature in Victoria taking place in the afternoon of the 9th, when winds were still westerly. The minimum at Adelaide fell to 36.8 on the night of the 9th, about the lowest on record there for October. Probably the real significance of the phenomenon lies in the fall of pressure which took place from the southwards between the 8th and the 10th, causing the 30.0 isobar which lay east and west from Cape Borda to Gabo Island on the 8th to turn as it were upon an axis so as to lie north and south less than 48 hours later. Over Northern Victoria the barometer fell from 30.1 to 29.7 within 24 hours from this cause, that is, to some impulse apparently coming from due south. The interesting question then arises, "To what extent was the cold and storm production generally due to rarefaction at the rain-producing levels?" The similarity between the phenomena of this storm and that of 28th September-1st October, 1908 (Figures 137-139) is worth noting. An examination of five other cases of unusual cold with snowfall in Melbourne shows this tilting of the rear isobars of the "low" causing them to lie north and south to be a common feature. This involves a displacement of the following "high" centre from a normal position over South Australia to one several degrees further south, and the "high" is always an intense one.

---

PROLONGED HOT SPELLS—SOUTH-EASTERN AUSTRALIA.

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

Figs. 131 to 136.

Hot Spells in Southern Australia.—In the southern parts of the continent or, say, within 200 or 300 miles of the south coast-line, spells of heat unpleasantly great are usually of short duration. The immediate cause is a low-pressure trough connected with an Antarctic disturbance, and as the average rate of summer Antarctics is about 800 miles per day the northerly winds in advance of the trough rarely last more than 24 hours. This is more expecially the case on or near the coast-line; the further inland we go the longer the preliminary period during which light variable or easterly winds preceding the true northerly induced by the trough are allowing an accumulation of heat. But occasionally even coastal towns are subjected to hot spells lasting several days. The prime factor for this is always a stationary anticyclone centred over Tasman Sea, but with its western slopes overlying eastern and south-eastern Australia. This usually goes with some slow-moving monsoonal depression over the western interior of the continent, or an Antarctic disturbance which fails in its advance eastwards. The real problem here is the cause of the stationary condition of the anticyclone. This set of maps provides reason for thinking that it may be the same as has already been noted in connexion with the dry winter of 1899, when cyclones off the east coast held back or built up the following anticyclones so as to form a semicircle of high pressures on their south-eastern and south-western to north-western boundaries. In this case there is evidence that a tropical cyclone for some days was following the usual parabolic course not near the east coast, but away to the north-east in the neighbourhood of New Caledonia. Such a course would certainly tend to maintain high pressure over Tasman Sea, and it may be that this is the usual cause of the phenomenon under review. The maximum temperatures in Melbourne for the five days 30th January—3rd February, 1912, were 96.4°, 102.6°, 106.5°, 105.9°, and 102.5° respectively.

Much the same operating factors appear to have been at work during the still more severe and, in Melbourne, record spell of heat from 15th to 20th January, 1908, the successive maxima at the Weather Bureau being 102.0°, 106.7°, 109.3°, 104.1°, 105.7°, and 107.5°. The charts for this period also show evidence of the existence of a very persistent tropical low-pressure system operating for some days away to the north-east beyond the high over Tasman Sea. This is most marked from the 16th to the 19th, when barometers were falling over Norfolk Island and the extreme north of New Zealand, and rising over eastern New South Wales and Victoria. On the 20th, when the tropical low was moving off, as shown by pressures rising again at Norfolk Island, the centre of the "high" moved northwards, allowing the trough of a slight Antarctic to move eastwards along the south coast of Victoria, thus bringing the much desired cool change. The charts for the 16th, 18th, and 20th illustrate the changes noticed on this occasion.

Figs. 137, 138, and 139.

East Coast Cyclone of Antarctic Origin.—The appearance on the coast of New South Wales of a cyclone owing its origin to the passage of an Antarctic disturbance is a somewhat irregular phenomenon, being of comparatively frequent occurrence in some years, or groups of years, and rare in others.^[1] The case here shown is a fairly typical one. On the 26th September, 1908, an ordinary Antarctic "Λ" depression lay over the Bight, while a high pressure system covered the eastern half of the continent. By the 28th, the low had passed Gabo and extended well up the New South Wales coast-line, the southerly winds in rear of it being very cold, with showers and squally weather. By the 30th, the Antarctic disturbance proper had passed completely away, but an elliptical depression was left over Tasman Sea, the coastal weather being still dominated by southerly winds, and very cold and wet. Snow fell extensively over the mountains in New South Wales. Next day the circulation was definitely cyclonic, and the weather warmer owing probably to all southern connexion being severed by high pressures forming to southward over Tasman Sea. The storm centre then moved off in a south-easterly direction.

Figs. 140 and 141.

East Coast Cyclone of Inland Origin.—A typical example of this happened on the 18th January, 1911. On the previous day a very pronounced monsoonal trough extended from Western Queensland into Victoria, the line of lowest pressures being slightly below 29.8 inches. Moderate to heavy rain fell during the next 24 hours in front of this line, and on the 18th, the eastward advance of the trough carried it over the New South Wales coastline, where pressures promptly fell some two-tenths of an inch and the air circulation became of the cyclonic type. The fall in temperature along the New South Wales coast produced by its rear circulation was very slight.

EAST COAST CYCLONE OF ANTARCTIC ORIGIN.

EAST COAST CYCLONE OF INLAND ORIGIN.

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

EAST COAST CYCLONE OF OCEANIC ORIGIN.	TROPICAL CYCLONE, NORTH QUEENSLAND.	DEVELOPMENT OF MONSOONAL TROUGH.

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

Figs. 142 and 143.

East Coast Cyclone of Tropical Oceanic Origin.—31st May—2nd June, 1910. This storm was first definitely indicated on the 31st May by falling barometers over New Caledonia and the North Queensland coast, with rain setting in rather extensively over the Central portions of Eastern Queensland. During the next 24 hours heavy rains fell on the coast south from Townsville, and moderate amounts were recorded generally over the south-eastern half of the State. On 2nd June, a definite cyclonic storm was centred off Brisbane immediately south of which some tremendously heavy rain fell. Cleveland had no less than 11.20 inches in the 24 hours. The winds though strong do not seem to have been of exceptional force. The storm continued its southward course for another day, bringing rain to the eastern parts of New South Wales, then moved off towards New Zealand. These storms may occur at any season, and are felt most severely along the coast-line between Brisbane and Sydney, where they develop a wind force and coastal rain production much beyond what their previous history while approaching oversea from the north or north-east would suggest. The steep barometric gradients induced on their south-west quadrants by anticyclonic conditions advancing eastwards over southern Australia probably assist to some extent.

Figs. 144 and 145.

Tropical Cyclone, North Queensland.—10th and 11th February, 1911. Storms of this type in North Queensland are confined in time almost entirely to the summer months, December to April inclusive. They usually strike the coast-line some few degrees further south than the position shown on the 10th, in which case they approach the coast from some point north of east preceded by overcast skies, heavy rain, falling barometers, and freshening winds from between S.S.E. and east. A heavy sea swell, the direction of which is a good index of the position of the storm centre, usually precedes the hurricane winds by some hours.

The storm shown here appears to have developed cyclonically from a depression with very slight gradients which lay over the Gulf of Carpentaria for two or three days previously. As the latitude is but little more than that usually assigned for the origin of tropical cyclones it is probable that the genesis of one is here represented. The chart of the 11th shows the centre further south, about Cardwell; next day it was just inland from Mackay with diminished energy, and it soon after lost its identity altogether. These cyclones usually develop winds of hurricane force between Cooktown and Townsville severe enough at times to wreck buildings. In this case, however, not much damage appears to have been done by wind, but very heavy rains fell—Port Douglas recording 11.88 inches in the 24 hours ending 9 a.m. on the 11th, and the rivers were flooded. These storms are seldom or never experienced with any severity on the New South Wales coastline.

Figs. 146 and 147.

Development of a Monsoonal Trough.—24th and 25th March, 1913. The way in which what appears to be an ordinary monsoonal dip in the isobars may under convectional action develop a barometric trough cutting the high to southward in two and joining the tropical and Antarctic low pressure belts is well shown here. Such an occurrence is practically always associated with rain, frequently heavy inland, and this, contrary to the usual experience inland, may be accompanied by southerly winds. The most reliable indications of rain in such cases are widespread cloud formation in the "dip" and a flow of upper air, as shown by the movements of cirrus or other high-level clouds over Victoria from some northerly point.

Figs. 148, 149, and 150.

Great Dust Storms.—11th-13th November, 1902. The raising of dust in Australia in summer time can hardly be said to require any special type of weather. That it be windy is usually sufficient. There are, however, seasons in which the production of dust becomes specially easy. Such was the spring of 1902, when the rainfall inland over large areas was so scanty that the surface soil once broken never became sufficiently compacted again by moisture to prevent strong winds, especially those of variable direction, from carrying it away. Many carefully worked fallowed lands in the northern parts of Victoria were practically swept bare of the loose soil mulch which had been so laboriously prepared by cultivation to conserve moisture and provide plant food for the next year's crop. Many a boundary fence was obliterated and lane filled to a depth of several feet by wind borne sand and loam during this year. In New South Wales the same occurred over large areas of purely pastoral country.

The charts for 11th, 12th, and 13th November, 1902, show all that barometer readings can tell us of the genesis and development of the worst of these dust raising storms. The morning of the 12th in Melbourne was beautifully fine, calm, and pleasant, the sky only partly covered by thin cirrus clouds moving from W.N.W. About 1 p.m. the calm was broken by a violent burst of northerly winds, which an hour or two later gave place to almost equally violent westerlies, and these to fierce north-westerly squalls a little later. The dust was at times suffocatingly dense, and the upper air was so loaded with it that the sun was rarely visible. In the country the effects were much more marked. At many inland towns the darkness produced almost equalled that of the blackest night, and in the houses nothing could be done without lamps or other means of lighting. Added to this were some phenomena of an even more terrifying character. At Boort and in some parts of the Riverina the storm was accompanied by a sort of globular lightning, "fireballs" were seen falling on the fields and roads, and scattering the earth. As these electrical phenomena were produced in a dry atmosphere the assumption is that they were due in some way to friction between dust particles.

The weather charts themselves are most interesting, and show that atmospheric changes and movements of a most unusual character were in progress. The chart of the 10th shows unusually quiet weather over the continent, the

DUST STORMS.

FOGS IN MELBOURNE.

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

only low-pressure system being an Antarctic passing eastward over and south from Tasmania, and a large monsoonal depression covering the western and northern interior of the continent. The two were separated by a large area of "flat" barometric pressures, never varying much from 30.0 inches. This is, of course, known to be rather favorable to the development of local squalls and thunderstorms inland, but these do not seem to have occurred, at all events in any isolated fashion, as no rain fell in New South Wales and very little in Victoria. On the morning of the 12th a "low," apparently cyclonic, was centred off Robe, where the barometer reading had fallen during the 24 hours from 29.99 to 29.52, nearly half-an-inch. The formation of this storm centre seems to be intimately related to the monsoonal depression of the preceding day, as a well-marked trough along the front of which northerly winds blow runs well beyond Alice Springs. These winds would necessarily be hot—at 9 a.m., 97° at Alice Springs; 95° at Farina; 90° at Broken Hill, &c.—and the rapid fall of pressure to southward producing steep gradients made them strong. Next day's chart shows the intensification still in progress. Gradients were very steep, especially in rear of the trough, the fall in pressure from Adelaide to Southern Tasmania being exactly 1 inch. Hence wind strength did not lessen at all with change of direction as the trough passed. All the conditions necessary to maximum dust production were, therefore, present—(1) preceding drought; (2) great heat; (3) strong and variable winds; (4) no rain with the storm itself.

Three months later, on the 14th February, 1903, a very similar storm occurred, dust being carried upwards to such an extent that the clouds of alto-cumulus level became of a curious copper colour, and the rain which afterwards fell was at first loaded with dirt. In many parts of Victoria remote from the inland plains the surface of the ground took on a different tint owing to the foreign matter deposited upon it. The amount of surface soil transported in this way during this summer was undoubtedly enormous. Another effect of these wind storms was a redistribution of plant life. In the Mallee the number of new varieties of grasses and small flowering plants springing up after the next winter's rains was a matter of frequent comment.

Figs. 151, 152, and 153.

Fogs.—Although fog is not usually regarded as one of the most important meteorological phenomena, its occurrence often has a very practical bearing upon the affairs of life. In the country, and especially among the hills, it may provide no inconsiderable portion of the moisture supply, while at other times its formation when destructive frosts are threatened is very welcome. In large cities its value is not so apparently positive, and from some aspects becomes very decidedly negative. This is seen in its effect upon the speed of traffic and consequent hindrance to business. What its value may be in providing excuses for tardy arrivals in public offices is, of course, another matter.

Fogs are very local phenomena, the general causes in the way of pressure, distribution, &c., which favour their development in one locality not being at all effective in others. One factor must, of course, be present in every case; that is the presence of water vapour in the lower stratum of air in sufficient quantity to produce saturation before the cooling by radiation or otherwise has reached its limit. Cooling by radiation being the most frequent cause of fog, it follows that the sky should be clear, and this most frequently happens during the passage of an anticyclone. But the air is then usually very dry, so much so that under ordinary circumstances inland in the winter months frost is the more probable result. The water vapour necessary for fog production is most likely to be present inland when a clear, calm night follows rain, and especially when the high-pressure system forms a wedge separating two "lows."

Fogs in Melbourne are most frequent in the early winter months, May—July. The conditions most favorable for their production are well illustrated by the three charts shown. It nearly always happens that the anticyclone is centred to southward. The effect of this is obvious. The gentle outflow of air from the high along the earth's surface reaches Melbourne from points between south and east, having just passed over a water surface, and so carrying a fair amount of moisture. The fact also that a cyclonic depression has very frequently been operative just before off the New South Wales coast and driven masses of humid air into Bass Strait would aid in this result. In connexion with this latter factor the influence of the mountains or hilly country stretching southwards from the Yarra sources into South Gippsland cannot be ignored, inasmuch as it causes the dissipation of the clouds of lower levels coming from the east or south-east, and thus aids in producing the required clear night sky over Melbourne. A very favorable type is that of 21st February, 1913. The tendency towards a tropical "dip" in the isobars seems to indicate sufficiently humid conditions inland. One of the densest fogs ever experienced in Melbourne was on the night of the 26th July, 1910. In this case heavy rains had just fallen over Gippsland, and the high is but the separation between two "lows." During very dense fogs the air movement is often from the north-east. This is not necessarily the direction of air movement which preceded the fog formation, but it helps to illustrate the usually small convectional action with light north—easterly breezes, as was noted previously in a paper dealing with observations upon the amount of dust ordinarily suspended in the atmosphere of Melbourne.

---

MARONG TORNADO. (Victoria.)	TORNADO-LIKE WINDSTORMS PRECEDING THE GEMVILLE TORNADO.	AVENEL TORNADO (?) (Victoria.)

Note.—Shading shows where rain fell during previous 24 hours. Wind direction shown thus ⟶ strong winds ⤜

Tornadoes.

Figs. 154 to 159.

The name "tornado" is given to an excessively violent and destructive wind storm affecting only a narrow strip of country, and producing its destructive effects, not by a straight blow, but by air in rapid rotation, as in a whirlwind. From the latter, which is relatively insignificant, the tornado differs essentially, as it does also from the vastly wider cyclone, though the terms are often confused.

The tornado is always associated with thunder and hail storms of extremely violent type. If we regard a thunderstorm as due to the uprushing of a column of air, from, say, the 4,000 to the 20,000 or even 30,000 feet level this giving rise to, and in turn being maintained by, the condensation of aqueous vapour, with the resulting phenomena of rain, hail, electric discharge, &c., and this column of air to take on a rapid spiral movement, which it naturally does, then the downward extension of this spiral movement to the ground provides the tornado. It is in respect to this thunderstorm origin that it differs from a whirlwind, the latter usually originating at ground level and not rising high enough to cause condensation in the very dry, hot air in which it occurs. The radius of action of the tornado may not be much greater than the few yards covered by an ordinary whirlwind, and rarely exceeds one-fourth of a mile, but what it lacks in area it more than makes up in intensity. From the cyclone it differs in the area affected, but, nevertheless, it is not to be regarded as a miniature cyclone. The tornado is a part of one thunderstorm; the cyclone is a vastly wider circulation of the air set in motion, at all events when of tropical origin, by the prevalence over a considerable area of the earth's surface of conditions which may be incidentally indicated by the occurrence of thunderstorms and even tornadoes in isolated parts of it. Some of the primary essentials to these conditions would be heat and atmospheric humidity above normal. It may be suggested, too, that the vertical temperature gradient would provide a means of definitely separating the two. In the tornado or thunderstorm the rising air must at any level, except possibly near the top, be warmer than the surrounding air at the same level, while above the cyclone the air soon becomes actually colder than at the same levels in the surrounding anticyclones. This is, of course, only another way of saying that a steep vertical temperature gradient is favorable for the occurrence, first of thunderstorms, and ultimately of cyclones.

Typical tornadoes are commonly supposed to be confined to North America. This is only true to the extent that they are undoubtedly more frequent and probably more violent there than elsewhere. Australian experience provides many genuine examples, but owing to the sparse population and the character of the storms themselves they have not yet been the subjects of very accurate scientific observation.

Judging by the records available, New South Wales and Victoria appear to be the States most liable to these visitations. In New South Wales they are most frequent in the summer, occurring only in connexion with monsoonal depressions; in Victoria they seem quite as liable to occur in connexion with strong Antarctic low-pressure systems, and the numbers do not therefore show the same marked preferences for the summer season.

The conditions most favorable for tornadoes inland and in the summer are (1) high temperatures; (2) considerable humidity; (3) very small barometric gradient, to which may be added a very probable fourth factor, unusually steep vertical temperature gradient. These seem to be most frequently provided by extensive but comparatively shallow monsoonal depressions, which favour a wide gentle air flow southerly from the tropical interior.

The weather charts for 25th to 29th September, 1911, show conditions typically favorable and abundantly justified by results, inasmuch as two tornadoes resulted. The first was in Victoria at Marong, near Bendigo, on the afternoon of the 27th September, 1911. This supplied all the characteristic features of a tornado, the long inverted cone depending from the dense blue-black thunder cloud, the narrow track, 5 to 12 chains wide and 12 miles in length, along which indescribable damage was wrought, and the accompanying violent thunder and hail storms. Fortunately a photograph of the storm cloud with its pendent funnel was secured by a gentleman 3 miles distant, and this has been reproduced, together with a full description of the storm, &c., in the 1911 September number of the Australian Monthly Weather Report. Two days later the same atmospheric conditions resulted in a similar storm in New South Wales at Cudal, between Forbes and Orange. The results in the latter case were not so serious, as the storm occurred in a sparsely populated area.

The charts given will repay inspection. That of the 25th shows a great valley-like depression extending from the Bight to the Northern Territory. while the axis of a high-pressure system lies north and south over Eastern New South Wales and Queensland. A wide direct southward drift of air in front of the trough is thus secured. The chart of the 27th shows the same features, the very slow eastward movement allowing of steady accumulations of heat and increasing humidity. By the 29th the depression was well over South-eastern New South Wales.

In connection with this disturbance no favorable feature appears to have been wanting. Slight though the barometric gradients were, they had complete control of the atmospheric circulation, the air in front of the trough from cumulus base to cirrus level flowing steadily from points near north. This is always favorable to thunderstorm development, and visual observation was sufficient to show that this case was no exception, the writer, who was in Northern Victoria at the time, making a special note of the abundance day after day of "towering" cumulus and cumulo-nimbus clouds.

From the 1st to the 5th November, 1910, a cyclonic depression with very slight gradients, but influencing enormous areas, was slowly passing from the north-west coast across the continent. During the whole of this period barometric conditions, ideal from a thunderstorm point of view, prevailed over the whole of Eastern Australia. They were, if anything, even more favorable than in the preceding example, and of much wider scope. In addition to heavy thunderstorm rains over the greater part of the interior the following exceptional phenomena were reported. At Hergott Springs (S.A.), in the driest part of the continent, a "cyclone" accompanied by a violent thunderstorm passed over the township at 5 p.m. on the 4th instant. This wrecked one house, unroofed others, and threatened loss of life. The barometer fell extraordinarily—to 28.40 inches—but the value of the reading is unknown. On the same afternoon at Richmond (Q.), 800 miles away, a "terrific cyclone" damaged most of the buildings, derailed railway trucks, and bent double iron telegraph poles. Two inches of rain fell.

On the preceding afternoon, near Laverton (W.A.), a thunder and hailstorm, giving half-an-inch of rain, was accompanied by a "hurricane."

Though violent, the foregoing are not absolutely identifiable as tornadoes, but the observer's description of a storm next day at Gemville, a place in the dry north-western interior of New South Wales, leaves little room for doubt. The 5th, he said, will long be remembered. One inch of rain fell in half an hour, hail stripped the ground bare, and a "cyclone," the spiral motion of which was a feature, cut a track through the scrub several miles long. The thunder and lightning were very severe.

The description of a local storm of extreme severity and destructiveness in North-eastern Victoria is of interest, and shows that tornadoes or very similar phenomena may occur under other than summer conditions of heat, humidity, and barometric pressure. About 3 a.m. on the 16th July, 1904, near Avenel, a "fierce wind" swept from N.W. to S.E. over a strip of country about 2 miles long and 30 yards wide. It completely wrecked a farmhouse which lay in its track, killing two of its inmates, carried large pieces of furniture for over half-a-mile, tore up trees by the roots and transported them for considerable distances, and demolished all of an orchard over which it passed. The storm was immediately preceded by vivid lightning and heavy peals of thunder.

The weather charts of the 15th and 16th July show that this occurred in connexion with the final stages of a very intense Antarctic low-pressure system which had been moving slowly eastwards along the south coast-line for five days previously, and suggest that it was coincident with a sudden rise in barometric pressure which ended the "Antarctic" and introduced anticyclonic conditions. The passage of a low-pressure trough usually coincides with a maximum of atmospheric instability, and in this case it seems probable that the exceptional energy developed was in some way connected with the arrival of the colder and denser air of the anticyclonic front—possibly by this causing forced ascent of the warmer and moister air of the low-pressure system. But whichever was cause or whichever effect, we know by experience that a sudden change from cyclonic to anticyclonic conditions, as shown by a large rise in barometers at the end of a lengthy period of Antarctic low pressures, is almost invariably associated with violent atmospheric disturbances and not infrequently thunder and hail storms.

---

1. The fact that East coast cyclonic storms of antarctic origin are, for long periods, rare, helps to emphasize the very abnormal character of the weather of the latter half of 1905, in which no less than eight of these storms occurred. This is one of the many instances of definite tendency for special weather types to recur in certain seasons, thus giving to at least some seasons a kind of individuality. As the latter half of 1905 was, in Melbourne, the coldest on record, two abnormalities at any rate are in agreement.