some storm-related weather trends in weymouth · some storm-related weather trends in weymouth...

Geoff Kirby BSc(London), BSc(Open)

6 College Lane, Weymouth, Dorset, DT4 7LP

� 01305-787253

[email protected]

Some Storm-Related Weather Trends in Weymouth DRAFT

10 December 2009

SUMMARY

A previous report (http://www.geoffkirby.co.uk/CoastalReport.pdf) examined flood risks in the vicinity of Weymouth Harbour caused by extreme tidal effects, low barometric pressures and high winds. This report examines primarily the flooding risks over a broader area surrounding Weymouth caused by high rainfall. This exploits the unique 127-year record from the Weymouth Weather Station. These data show that rainfall trends are not conforming to those predicted by climate modelling for the Southern England region. Indeed, rainfall trends in Weymouth are cyclical with no clear long-term trends. The cycles have periods between 10 and 80 years. In general, the rainfall patterns being experience in this century were also experienced in the 20

th century.

As an example, whilst the annual rainfall has shown no significant change since 1881, the number of wet days each year is now about fifty days higher than the period 1940 - 1995. Weymouth is becoming wetter; the same annual rainfall is falling on a much greater number of days compared with the second half of the 20

th century. However, the present number of

annual wet days is much the same as experienced in the 1930s.

An analysis of wind statistics allowed a simple empirical equation to be proposed for predicting extreme wind frequency.

Some Storm-Related Weather Trends in Weymouth - Geoff Kirby

Draft

CONTENTS

Summary Cover

1.

2.

3.

4.

5.

6.

7.

Introduction

Rain records analysed

a. Climate change forecasts

b. Long-term rainfall

c. Decadal rainfall

d. Seasonal rainfall

Rain and barometric pressure

Wind

Conclusions

Acknowledgements

References and notes

Page 2

Page 7

Page 7

Page 9

Page 13

Page 17

Page 25

Page 32

Page 36

Page 37

Page 38


Draft

1. Introduction

In Part 1 of this series of reports 1 a mathematical model was generated to enable sea levels

and wave heights to be forecast for Weymouth Harbour. This model enabled predictions to

be made for the frequency of a sea defence wall being overtopped either by a consolidated

flow of water when the sea level is higher than the wall or by ‘slopping’ when the wave crests

exceed the wall height but the underlying sea level is below wall height.

The report concentrated on flooding of the Town Centre by water overtopping the existing or

a future sea wall. However, in the 20th

century the great majority of floods in the Weymouth

peninsula occurred in the Park District and were not primarily due to sea levels being

excessive.

All of the floods shown in figure 2 (overleaf) were due primarily to excessive rainwater causing

the River Wey to burst its banks. In some cases the river flooded the town because high tides

in Weymouth Harbour prevented the sluice gates opening in Westham Bridge. However, this

was not always the case.

Figure 1 - The Weymouth peninsular


Draft

Figure 2 - A collage of 20th

century floods in Weymouth


Draft

This second report examines trends in eighty-one years of Weymouth’s daily weather records

and studies the relationships between rainfall and other meteorological parameters such as

barometric pressure and wind strength.

This report also examines the ‘return frequency’ of extreme weather events which potentially

can cause flooding. After a flood there are always calls for greater protection to be installed -

and rightly so. However, the cost of installing protection has to be balanced against the

frequency with which such events occur and how long the useful life of such a protection

scheme might be.

By way of illustration consider an extreme case of the hurricane of 1824.

This raised waves on 23 November 1824 estimated at tens of metres in height which swept at

high speed over Chesil Bank and inundated fields well inland from the northern shoreline of

The Fleet. Fleet Church was largely destroyed and many local people were left homeless.

Melcombe Regis 2 Esplanade was totally destroyed and the town was extensively inundated.

Many inhabitants died as the sea poured over the beach and ran through to the Backwater.

About 50 people were killed at Chesil on Portland and 80 cottages were destroyed 3.

Now imagine a similar event occurring in 2010. Weymouth’s Esplanade and Harbour are still

vulnerable to severe storm damage and there would be very extensive flooding of the Town

Centre and the Park District.

One reaction in 2010 would be to blame global warming; a modern media-driven but baseless

diagnosis.

There would be an outcry against coastal planners for not having provided protection against

this freak hurricane and there would be pressure to build expensive protection in anticipation

of the next such event.

And yet, a storm of the severity of 1824 has occurred only once in recorded history of the

area and so providing protection against a repeat of this event would seem hard to justify.

The aftermath of the extensive damage caused by Hurricane Katrina to New Orleans in 2005

shows that there is also a political dimension to decisions about the provision of expensive

protection of communities as well as engineering and scientific considerations.


Draft

In this report Weymouth’s historical weather database is examined for trends in rainfall, wind

speed and barometric pressure; all of which can either directly cause flooding or can

exacerbate conditions under which flooding could occur.

It is hoped that this report will contribute to the scientific decision-making process.


Draft

2. Rain records analysed

a. Climate change forecasts

Climate modelling experts predict changing rainfall patterns in southern UK as a result of

climate change. The technique used is to run different models under a range of input

assumptions and then pool the results to determine a mean and spread of possible changes.

The estimates of precipitation change for the South West England are summarise below 4.

These have been derived using the ‘Weather Generator’ program 5.

Change in mean winter precipitation

2020 2050 2080

Low Medium High Low Medium High Low Medium High

+6% +7% +6% +12% +17% +18% +19% +23% +31%

Change in mean summer precipitation

2020 2050 2080

Low Medium High Low Medium High Low Medium High

-7% -8% -5% -14% -20% -20% -12% -24% -30%

Table 1 - Predicted mean change in precipitation

The changes in precipitation are relative to the mean for 1961 - 1990.

The table covers three assumed greenhouse gas emissions achievements, ‘Low’, ‘Medium’

and ‘High’. These are defined in terms of many factors including GDP growth, population

growth, land use change, future aircraft use, achievement of energy efficiency targets, etc 6.

The above figures are the medians lying within large bands of uncertainty. For example, the

figure of +7% highlighted in red is bounded by lower and upper decile figures of -2% and

+20% respectively.

Evidence to support these predictions is weak. For example, we read 7 8

“There has been an increase in average winter precipitation in all regions of the

UK between 1961 and 2006. However this trend is only statistically significant

above background natural variation in Northern England and Scotland where

increases of 30 to 65% have been experienced.”


Draft

“There has been a slight decrease in average summer precipitation in most

regions of the UK between 1961 and 2006. However this trend is not statistically

significant above background natural variation.”

“There are no statistically significant trends in the average number of rain days

or mean sea level air pressure for any region of the UK between 1961 and

2006.”

We see that the lack of supporting evidence is due, in part, to a lack of statistically meaningful

data trends. The problem may well lie with the relatively short set of daily weather records

collected at Ringway Airport between 1961 and 1990 used to calibrate the ‘Weather

Generator’ program 9. This is where the Weymouth daily weather database has the

advantage having been started in 1927 10

. This not only has a better long-term statistical

significance but applies locally. It is well known that Weymouth weather is sometimes very

different from Dorchester which is only 12 km to the north behind the Ridgeway Hills which

significantly change the weather pattern between the two towns.

In general, the predictions for the south west of England are for wetter winters and drier

summers with more days having intense rainfall in the winter than now.

What can Weymouth’s weather records tell us?


Draft

b. Long-term rainfall

0

200

400

600

800

1000

1200

1400

1600

1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Year

Av

era

ge

Ra

infa

ll (

mm

/yr)

From daily records

From annual records

Annual Rainfall

Figure 3 - Annual rainfall records

In figure 3 the annual mean rainfall has been plotted using annual summaries before 1927

and the average of daily measurements after that date.

The figure tends to show no overall change in annual rainfall since 1881 although there are

short-term variations from year to year which may be no more than random variations not

associated with any underlying physical mechanism.

Table 2 shows the probability of observing rainfall on any day in the period 1926 - 2008

exceeding the amount shown stated.

No Rain 0.612 >18 mm 0.0124

>0 mm 0.388 >20 mm 0.0066

>2 mm 0.228 >30 mm 0.00270

>4 mm 0.160 >40 mm 0.00076

>6 mm 0.115 >50 mm 0.00027

>8 mm 0.082 >60 mm 0.0000995

>10 mm 0.056 >70 mm 0.0000995

>12 mm 0.042 >80 mm 0.0000663

>14 mm 0.0300 >90 mm 0.0000332

>16 mm 0.0225 >100 mm 0.0000332

Table 2 - Probability of observing rainfall exceeding the stated amount on any day,


Draft

Figure 4 shows data in table 2 plotted and it can be seen that the cumulative probability

distribution is close to an exponential curve until a rainfall exceeding about 50 mm after which

the probability of observing large amounts of rainfall in a day remains much the same.

The extreme right-hand data point is for 18 July 1955 when 182 mm of rain fell in Weymouth

causing extensive flooding in the Town Centre and the Park District 11

.

Distribution of Daily Rainfall (1926-2008)

0.00001

0.0001

0.001

0.01

0.1

1

0 20 40 60 80 100 120 140 160 180 200

Daily Rainfall (mm)

Cu

mu

lati

ve P

rob

ab

ilit

y

Figure 4 - Probability of observing rainfall on any day exceeding stated value

It may be surprising to see that rain has fallen on nearly 40% of days over the past eighty-two

years. The impression that days are drier than this may well arise because we are mostly

aware of rain in the daytime rather than rain that falls only overnight. This analysis does not

distinguish between the distribution of rainfall during 24-hourly periods.

Figure 5 shows the probability of observing rainfall on any day over the same eighty-two year

period.


Draft

Probability Distribution of Daily Rainfall (1926 - 2008)

0.0001

0.001

0.01

0.1

1

0 2 4 6 8 10 12 14 16 20 24 30 40 50 60Daily Rainfall (mm)

Pro

bab

ilit

y o

f O

bserv

ing

Rain

fall

Figure 5 - Probability of observing rainfall on any day within stated values

Date mm Date mm Date mm

18/07/1955 182.0 01/12/2005 41.1 27/11/1929 35.3

15/07/1937 83.0 11/07/1953 41.0 26/01/1940 35.0

11/07/1977 78.0 04/11/1966 41.0 12/01/1948 35.0

22/10/1966 57.0 13/10/1939 40.0 21/01/1962 35.0

05/06/1983 56.0 22/10/2003 39.8 25/09/1967 35.0

18/09/1999 53.7 08/10/1988 39.0 30/05/1979 35.0

18/10/1955 51.0 24/02/1933 38.9 23/10/2005 34.4

05/08/1997 50.6 25/05/1941 38.0 06/03/1941 34.0

02/07/1957 50.0 15/10/1966 38.0 08/10/1960 34.0

28/09/1991 50.0 13/11/1974 38.0 18/03/1964 34.0

21/09/1949 49.0 21/09/1976 38.0 06/08/1966 34.0

02/07/1950 46.0 31/07/1978 38.0 03/08/1974 34.0

25/07/1954 46.0 13/06/1980 38.0 29/11/1985 34.0

16/11/1935 44.1 25/06/2006 37.9 31/08/1988 34.0

04/09/1974 44.0 07/11/1926 37.0 09/08/1989 34.0

13/11/1940 43.0 18/06/1957 37.0 13/12/1989 34.0

10/08/1960 42.0 06/11/1969 37.0 30/12/1993 34.0

27/12/1979 42.0 09/09/2002 36.1 06/09/1927 33.3

14/09/1927 41.5 01/10/1927 36.0 29/10/1937 33.0

27/12/1928 41.5 22/06/1983 36.0 29/12/1955 33.0

Table 3 - The dates and rainfall for the sixty wettest days in Weymouth (1926 - 2008)

It is curious to note that six out of the top thirteen wettest days fell in July.


Draft

Figure 6 - Average number of days each year delivering more than a specified rainfall.

Figure 6 shows the number of days each year from 1927 to 2004 upon which rain fell in

Weymouth.

It can be seen that the number of wet days was about 50 days higher than average in the

1920s and 1930s and again around 1995 - 2004 although the current trend is for Weymouth's

wet days to be about 20 more than the long-term average.

There is no evidence that Weymouth's long-term trend is for more wet days each year.

Indeed, since 2000 the number of wet days has been generally falling.

However, the number of days each year when rain falls has increased by about 50 days since

1990 but the number of days on which it rains hard has not changed. This means that there

are more days with light rain.

However, this increase in the number of wet days until 1990 follows a prolonged period of

decreasing number of wet days since about 1940. The figures for wet days are now about the

same as in the late 1920s and 1930 strongly suggesting that global warming is not the cause

of these changes but, rather that they are due to a long-term natural cycle.

The long drought of 1976 shows up quite clearly in figure 6.


Draft

c. Decadal rainfall

Days per Decade with Rainfall

0

200

400

600

800

1000

1200

1400

1600

1800

2000

2200

1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Nu

mb

er

of

Da

ys

pe

r D

ec

ad

e

Scaled from 9 years

Figure 7 - Number of wet days each decade

Figure 7 shows the number of days each decade on which measurable rainfall was observed.

The figure for the 2000s covers only nine years and so the data have been scaled up.

It can be seen that there is a modest trend for the number of wet days per decade to increase

from the 1940s after a drop from the wet decade of the 1930s. There is a hint of a cyclical

behaviour with a period of 30 - 40 years superimposed upon an underlying increase. This is

highly speculative however.


Draft

Days per Decade with Rainfall Exceeding 10 mm

0

50

100

150

200

250

300

1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Nu

mb

er

of

Da

ys

pe

r D

ec

ad

e

Scaled from 9 years

Figure 8 - Number of days each decade with rainfall exceeding 10 mm.

Figure 8 shows the number of days each decade on which at least 10 mm of rain fell. There

is no clear trend in these data.


0

20

40

60

80

1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Nu

mb

er

of

Da

ys

pe

r D

eca

de Scaled from 9 years


Figure 9 shows the number of days recording more than 20 mm of rain each decade. There

appears to be a distinct trend for a regular and possibly cyclical variation having a period of


Draft

about six decades. If this is a true interpretation of the data then the 2010s could deliver a

50% increase in the number of days producing at least 20 mm of rain. This will have a

significant effect on flooding of the River Wey catchment and potential flooding area.


0

5

10

15

20

1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Nu

mb

er

of

Da

ys

pe

r D

eca

de

Scaled from 9 years


Figure 10 shows the number of days recording more than 30 mm of rain each decade. There

appears again to be a distinct trend for a regular and possibly cyclical variation having a

period of about six or seven decades. However, the number of days in the sample is too small

to be statistically significant.

If the apparent trend is valid this predicts that the number of very wet days will increase

throughout the 2010s and 2020s to reach double the figures observed in the 1990s and

2000s.


Draft


0

1

2

3

4

5

6

7

1930s 1940s 1950s 1960s 1970s 1980s 1990s 2000s

Nu

mb

er

of D

ay

s p

er

Dec

ad

e

Scaled from 9 years


Figure 11 shows the number of days recording more than 40 mm of rain each decade. It is

obvious that the 1950s will be remembered for torrential downfalls and, indeed, the Park

District experienced some of its worst floods in that decade.

Overall the number of days in the sample is too small to allow trends to be derived.


Draft

d. Seasonal rainfall

Figure 12 shows the mean quarterly rainfall from 1927 to 2008 inclusive.

More rain falls in the period October - December inclusive than in other quarters.

Also shown on this chart are the Standard Deviation (SD) figures for each quarter.

Average Rainfall by Quarter

0

50

100

150

200

250

300

1 2 3 4Quarter

Ra

infa

ll (

mm

)

Mean

SD

Figure 12 - Average quarterly rainfall

It is interesting to predict the rainfall for the rest of the 21st century combining the above chart

with the forecasts on table 1.


Draft

Predicted Rainfall Trends - Southern England

Low Emissions Scenario

0

200

400

600

800

1000

1980 2000 2020 2040 2060 2080 2100 2120

Year

Ra

infa

ll (

mm

)Winter

Summer

Annual

Figure 13 - Predicted rainfall trends for Southern England


Medium Emissions Scenario

0

200

400

600

800

1000

1980 2000 2020 2040 2060 2080 2100 2120

Year

Ra

infa

ll (

mm

)

Winter

Summer

Annual



Draft


High Emissions Scenario

0

200

400

600

800

1000

1980 2000 2020 2040 2060 2080 2100 2120

Year

Ra

infa

ll (

mm

)Winter

Summer

Annual


We see from the above three charts that the rainfall is predicted to reduce in the summer

months and increase in the winter; there being a modest rise in overall rainfall (up to 6%) for

all three emissions scenarios by the end of the 21st century.

Total Rainfall in 1st Quarter

0

50

100

150

200

250

300

350

400

1920 1940 1960 1980 2000 2020Year

To

tal

Ra

infa

ll (

mm

)

Figure 16 - Rainfall in the first quarter of each year.


Draft

Figure 16 shows the annual rainfall (mm) in the first quarter of each year and it can be seen

from the five-year running average (the blue curve) that there has been no significant change

in rainfall patterns over the eighty-one year period.

The ‘Weather Generator’ predictions for winter rainfall in Southern England are for increases

per century of 22%, 27% and 41% respectively for ‘Low’, ‘Medium’ and ‘High’ greenhouse gas

emissions over the 21st century.

If anything, the trend in figure 16 shows a general decrease in rainfall in the first quarter of the

year with no sign of the forecast increase.

It is clear from figure 16 that there is a cyclic variation in mean rainfall with a period of about

10 - 11 years which is close to the mean period of sunspot cycles which is 10.7 years 12

.

There is a long history stretching back to Norman Lockyer in 1868 purporting to show that

there are links between sunspot cycles and weather 13

.

The cases for and against an influence of sunspots and their cycles on weather will no doubt

rage on for decades 14

15

16

17

18

.

Global Land+Ocean Surface Temperature Anomaly (C) (Base: 1951-1980)

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1860 1880 1900 1920 1940 1960 1980 2000 2020

Figure 17 - Global Mean Temperature


Draft

Figure 18 - Global Mean Carbon Dioxide concentration 19

However, the lack of a strong correlation between Global Mean Temperature and CO2

concentrations - see figures 17 and 18 above - shows that the latter is not the only

mechanism actively driving climate change. Many mechanisms have been proposed

including:

- Volcanic sulphurous particles ejected into the atmosphere 20

,

- The post-WW2 increased use of leaded fuel resulting in increased aerosol

concentrations 21

,

- Decadal variations in the North Atlantic Oscillation ocean flows 22

,

- Sunspot variations 23

.

In order to investigate the possible influence of sunspot cycles on the First Quarter rainfall records, the dates when each recent sunspot cycle started

24 have been superposed onto

figure 18 as shown below.


Draft

Total Rainfall in 1st Quarter

0

50

100

150

200

250

300

350

400

1920 1940 1960 1980 2000 2020Year

To

tal

Ra

infa

ll (

mm

)

Figure 19 - Sunspot cycles superimposed onto figure 18

When the actual sunspot cycles are superimposed onto the data we see an initially good correlation over the period 1927 - 1985 but this falls apart after 1985.

It should be clear that sunspot cycles are not, after all, a factor in Weymouth’s rainfall.

Total Rainfall in 2nd Quarter

0

50

100

150

200

250

300

1920 1940 1960 1980 2000 2020Year

To

tal

Ra

infa

ll (

mm

)

Figure 20 - Rainfall in the second quarter of each year.


Draft

Figure 20 shows the annual rainfall (mm) in the second quarter of each year and it can be

seen from the five-year running average (the blue curve) that there has been no significant

increase since 1927 although there is a possible cyclical pattern with a period of about 17

years.

Total Rainfall in 3rd Quarter

0

50

100

150

200

250

300

350

1920 1940 1960 1980 2000 2020Year

To

tal

Ra

infa

ll (

mm

)

Figure 21 - Rainfall in the third quarter of each year.

Figure 21 shows the annual rainfall (mm) in the third quarter of each year and it can be seen from the five-year running average (the blue curve) that there has been a generally reducing rainfall since about 1960.

However, the third quarter mean rainfall in the 1990s and 2000s was about the same as in the late 1930s suggesting a long-term cycle in the weather.

It is interesting to see that the spectacular daily rainfall of 182 mm in July 1955 does not give a dominant contribution to that quarter's result as shown by the green circle on figure 21. In fact, the UK record-breaking rainfall in July 1955 appears to have been the culmination of fifteen years of increasing third-quarter rainfall after which the rainfall eased off.


Draft

Total Rainfall in 4th Quarter

0

50

100

150

200

250

300

350

400

450

500

1920 1940 1960 1980 2000 2020Year

To

tal

Ra

infa

ll (

mm

)

Figure 22 - Rainfall in the fourth quarter of each year.

Figure 22 shows the annual rainfall (mm) in the fourth quarter of each year and it can be seen from the five-year running average (the blue curve) that there has been a small but scarcely significant decrease from 1927 to 1980 with a possible small rise since then.


Draft

3. Rain and Barometric Pressure

It is traditional weather lore that low barometric pressure increases the likelihood of rain. This

is the assumption upon which barometer dials are designed as shown below.

Figure 23 - Traditional barometer dial

Every day since 1927 measurements and visual observations have been made of twenty-six

weather-related parameters at the Weymouth Weather Station.

In this section the database is used to examine the relationship between rain and barometric

pressure.

The importance of any correlation between rainfall and barometric pressure lies in the fact

that they both have a significant effect on the probability of flooding in Weymouth.


Draft

Effect of Barometric Pressure

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

970 980 990 1000 1010 1020 1030 1040P (mB)

∆∆ ∆∆H

(P)

(m)

Figure 24 - The relationship between barometric pressure and sea level.

Figure 24 shows the sea level height excess above the predicted astronomical tidal height

plotted against barometric pressure as derived in Part 1 of this series of reports 25

. It can be

seen that a reducing barometric height causes the sea level to rise. The regression equation

predicts an increase of 12.2 mm/mB which is close to the value corresponding to hydrostatic

rise in sea water.

Thus, if a pressure of (say) 990 mB occurs the sea will, on average, be about 300 mm higher

than predicted. This is a significant amount in Weymouth Harbour 26

where the tidal variations

are relatively small; mean Spring Tides varying between 0.2 m at low tides and 1.4 m at high

tides 27

.

A sea level rise of 300 mm can significantly increase the threat of flooding. If this coincides

with heavy rainfall the River Wey may not be able to discharge through the sluices at

Westham Bridge and the rain water could backup and flood from Radipole Lake into the

town; mainly into the Park District but also into the Town Centre if conditions are unfavourable

in Weymouth Harbour.


Draft

Approximately 30,000 daily weather records have been analysed to determine the

relationship between barometric pressure and rainfall.

That there is no deterministic relationship between rainfall and barometric pressure will first

be illustrated by referring to the period encompassing the spectacular rainfall and flooding on

18 July 1955.

Barometric Data for 1955

990

1000

1010

1020

1030

1040

29/05/55 08/06/55 18/06/55 28/06/55 08/07/55 18/07/55 28/07/55 07/08/55

Date

Baro

metr

ic P

ressu

re (

mB

)

Figure 25 - barometric records for Summer 1955 in Weymouth

Figure 25 shows the barometric pressure recorded at the Weymouth Weather Station in the

summer of 1955. These records are not unusual and the pressure is generally above the

mean of 1017 mB. Rain fell on the days indicated by the vertical red arrows and there is a

broad correlation between dips in pressure and rainfall. However, on 18 July 1955 as

indicated by the blue arrow, 182 mm of rain fell on Weymouth and 280 mm fell at Martinstown

28.

And yet, there is no indication from the barometric record that such vast quantities of rain

were expected. In fact, at the time of the disastrous rainfall and consequent flooding the

barometric pressure was slightly above the mean of 1017 mB.


Draft

This rainfall caused exceptional flooding with a great deal of damage especially in

Martinstown and along the River Wey Valley all the way to Weymouth Harbour. Some of the

photographs in figure 2 were taken during this event.

There is, however, a broad statistical relationship between rainfall and barometric pressure.

Incidence of Wet Days

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

970

975

980

985

990

995

1000

1005

1010

1015

1020

1025

1030

1035

Barometric Pressure Range (mB)

Fre

qu

en

cy o

f W

et

days

Figure 26 - Incidence of wet days as a function of barometric pressure.

Figure 26 shows the probability that a day will experience precipitation as a function of

barometric pressure. Clearly, the lower the barometric pressure the greater probability that a

day will be wet. This should come as no surprise to even the most casual of weather

observers.


Draft

Quantity of Rain Fall

0

1

2

3

4

5

6

7

8

97

0

97

5

98

0

98

5

99

0

99

5

10

00

10

05

10

10

10

15

10

20

10

25

10

30

10

35

Barometric Pressure (mB)

Ra

in f

all

(m

m)

Figure 27 - Average rainfall on days that are wet as a function of barometric pressure.

Figure 27 shows the average rainfall on the wet days as a function of barometric pressure.

This chart shows that when barometric pressure is low the wet days deliver more rain than

wet days having high barometric pressure. Again this accords with everyday observation.

However, the combination of low barometric pressure, frequent wet days and each wet day

delivering more rain combine to increase flooding risks in Weymouth.

Sea Level Rise and Rainfall

0

1

2

3

4

5

6

7

-0.2 0 0.2 0.4 0.6Sea Level Rise in Weymouth Harbour (m)

Mean

Rain

fall

Per

Day (

mm

)

Decreasing Barometric

Pressure


Draft

Figure 28 - Mean effect of barometric pressure on rainfall and sea level rise in Weymouth

Harbour.

Figure 28 shows how the mean rainfall on any day is linked to mean sea level rise due to

barometric pressure changes. It can be seen that very low levels of barometric pressure can

lift the sea level by over half a metre whilst threatening as much as seven millimetres of rain

in any day.

These are of course mean values and there will be much variation about the values plotted.


Draft

4. Wind

There is a problem with Weymouth’s Weather Station records for wind.

The wind speed data appears to have been affected by the changing location of the recording

equipment and it may be unsafe to draw long-term conclusions.

0

2

4

6

8

10

12

14

16

1950 1960 1970 1980 1990 2000 2010Year

Av

era

ge

Win

d S

pe

ed

(m

ph

)

Figure 29 - Winds speeds in Weymouth

Figure 29 shows the wind speed record from 1949 averaged over a one-year moving window.

Before 1949 wind speeds were recorded as Beaufort Numbers which cannot accurately be

related to actual wind speeds.

It can be seen that there is an overall downwards trend in wind speed with a dramatic drop

between 1969 and 1992 as indicated by the red arrows. This change is almost certainly due

to the change in location of the recording equipment in these years. In 1969 the recorders

were moved from the Corporation Yard on Westwey Road to Westhaven Hospital. The latter

is near the crest of a hill on the outskirts of the town whereas the Corporation Yard was in the

centre of the town close to the Inner harbour and surrounded by buildings. One might expect


Draft

the winds to be stronger on the later, more exposed site. However, the above chart shows the

winds recorded as being about 4 mph lower.

In 1982 and 1983 the recorder was moved within the Westhaven Hospital grounds and this

may account for the sudden rise in mean wind speed seen in the chart at around that time.

In 1992 the equipment was moved because of vandalism which had been a big problem

since the 1950s. The current location is a secret but is significantly different from the

Westhaven site. This move almost certainly accounts for the jump in mean annual wind

speed recorded in 1992.

Various sources predict that the climate predictions also indicate that Britain will be windier.

One study suggests 30 percent more gales in Wales and Southern England in winter,

increasing the risk of another storm like that in 1987, which left £2 billion of damage in its

wake.

So, is it getting more windy in Weymouth as climate modellers are predicting?

Despite the discontinuities in the data caused by location changes, there do seem to be two

trends in operation in figure 29.

Firstly, there is an overall trend downwards of about 2 - 3 mph per decade which is seen in

periods where the recorder was in the same location, i.e., the 1970 - 1979 and the 1995 -

2005 decades.

Secondly, there seem to be peaks in the data approximately every 10 - 11 years as indicated

by the blue arrows.

In the rest of this analysis only wind data for 1993 - 2008 inclusive will be used when the

weather station was at its present site.


Draft

Smoothed Wind Speed

0

5

10

15

1990 1995 2000 2005 2010Year

Win

d S

pe

ed

(k

ts)

Figure 30 - Wind speed smoothed over running half-year window.

Figure 30 shows that, for a self-consistent run of measurements, there is a downwards trend

in the mean wind speed from 1993 to 2007 after which the trend recovers to the 1996 value.

The annual winter high wind effect is superimposed upon this trend.

There is no evidence that the predicted increase in wind speeds due to climate change is yet

occurring in Weymouth.


Draft

Probability of Wind Exceeding Stated Value

0.0001

0.0010

0.0100

0.1000

1.0000

0 10 20 30 40 50Wind Speed (kt)

Pro

ba

bilit

yProb = 3.0*exp(-0.2W)

Figure 31 - Probability of observing greater than stated wind speed

Figure 31 shows the probability of observing a wind speed greater than shown. The error bars

indicate that the number of data values is small towards the right-hand of the chart with only

three samples in the last point plotted.

Conditional upon there being no significant trend in wind speed distribution over longer

periods of time than used in the analysis (5,844 daily values) then the empirical curve shown

on figure 31 may be used to extrapolate to higher speeds observed over a longer period of

time.


Draft

Wind Speed and Barometric Pressure

0

1

2

3

4

5

6

7

975 980 985 990 995 1000 1005 1010 1015 1020 1025

Barometric Pressure (mB)

Ave

rag

e W

ind

Sp

ee

d (k

t)

Figure 32 - The relationship between barometric pressure and average wind speed

Figure 32 shows the relationship between barometric pressure and average wind speed. This

shows that the strongest wind speeds are observed, on average, when the barometric

pressure is around 990 mB with generally lower wind speeds at lower and higher pressures.

This may be because the strongest winds in a cyclonic or anti-cyclonic circulation are not at

the centre of the ‘eye’ of the storm where the pressure is lowest but somewhere between the

centre and the outer edge.


Draft

5. Conclusions

A previous report (http://www.geoffkirby.co.uk/CoastalReport.pdf) examined flood risks in the vicinity of Weymouth Harbour caused by extreme tidal effects, low barometric pressures and high winds.

This report examines primarily the flooding risks over a broader area surrounding Weymouth caused by high rainfall. This exploits the unique 127-year record from the Weymouth Weather Station.

These data show that rainfall trends are not conforming to those predicted by climate modelling for the Southern England region. Indeed, rainfall trends in Weymouth are cyclical with no clear long-term trends. The cycles have periods between 10 and 80 years. In general, the rainfall patterns being experience in this century were also experienced in the 20

th

century.

As an example, whilst the annual rainfall has shown no significant change since 1881, the number of wet days each year is now about fifty days higher than the period 1940 - 1995.

Weymouth is becoming wetter; the same annual rainfall is falling on a much greater number of days compared with the second half of the 20

th century. However, the present number of

annual wet days is much the same as experienced in the 1930s.

An analysis of wind statistics allowed a simple empirical equation to be proposed for predicting extreme wind frequency.


Draft

6. Acknowledgements

Ariel photographs are copyright Dorset County Council 2000 and are reproduced here with

permission.

Figure 2 is compiled from photographs in the Dorset County Library Archive and are

reproduced with permission.

The Weymouth weather data has been compiled as a result of the dedicated and untiring

work of a succession of volunteers. Bob Poots is the current weather observer. He very

generously made the data available to me in a suitable machine readable format.

Last and certainly not least I acknowledge the never failing patience of Sandra - seen on the

front cover of this report - who lives with a very old and eccentrically obsessive scientist. Her

untiring and uncritical support makes life worth living.

Biographical Notes. Geoff Kirby graduated in 1960 from London University with a First Class

Honours degree in Physics. He spent most of his working life at Portland with the Ministry of

Defence. In 1992 he took early retirement having been Head of the Oceanographic and

Sonar Performance Department for nine years. He then worked as a sole trading consultant

to a variety of companies as well as the MoD until finally retiring in 2004. In 2000 he

embarked on an Open University BSc degree course in Environmental Sciences with an

additional year studying the History of Mathematics. He graduated just before his 67th

birthday.


Draft

7. References and notes

1 Sea Level And Flood Risk Forecasts For Weymouth Harbour Geoff Kirby

(25 November 2009) available for download at

www.geoffkirby.co.uk/CoastalReport.doc or www.geoffkirby.co.uk/CoastalReport.pdf

2 What is now commonly named ‘Weymouth’ is actually Melcombe Regis. Historically

Weymouth was the area to the south of the harbour. Both areas will be referred to as

‘Weymouth’ here.

3 http://www.soton.ac.uk/~imw/chestorm.htm (accessed 02 December 2009)

4 http://ukclimateprojections.defra.gov.uk/content/view/982/527 (accessed 02 December

2009)

5 http://ukclimateprojections.defra.gov.uk/images/stories/UKCP09_WGenerator.pdf

(accessed 03 December 2009)

6 http://www.grida.no/publications/other/ipcc_sr/?src=/Climate/ipcc/emission/091.htm


7 Environment Agency (2002) How your region might be affected in the 2050s .

http://www.metoffice.gov.uk/climatechange/guide/ukcp/map/ (accessed 03 December

2009)

8 http://ukclimateprojections.defra.gov.uk/content/view/512/9/ (accessed 02 December

2009)

9 http://ukclimateprojections.defra.gov.uk/images/stories/UKCP09_WGenerator.pdf


10 In fact, weather measurements were made from 1881 but these are only available as

annual summaries.

11 About 280 mm fell in the Martinstown area about 10 kms north west of Weymouth.

This was a UK rainfall record for 54 years until November 2009 when slightly more rain

fell in Cumbria.


Draft

12 http://en.wikipedia.org/wiki/Solar_cycle (accessed 02 December 2009)

13 http://projects.exeter.ac.uk/nlo/about/nlockyer.htm (accessed 07 December 2009)

14 You can’t control the climate, Phillip Stott, New Scientist 20 September 2003 p 25

15 The Chilling Stars H. Svensmark and N. Calder, Icon Books (2008) ISBN 978-

1840468-66-3

16 Sunspots are up, here comes the rain New Scientist 8 November 2008 p10

17 Saved by the sun New Scientist, 16 September 2006, p32

18 A fake fight A. Thorpe New Scientist 17 March 2007 p 24

19 http://en.wikipedia.org/wiki/File:Mauna_Loa_Carbon_Dioxide-en.svg (accessed

02 December 2009)

20 http://en.wikipedia.org/wiki/Global_dimming#Probable_causes (accessed 24

November 2009)

21 http://www.newscientist.com/article/dn16976-did-lead-cause-global-cooling.html

(accessed 24 November 2009)

22 http://www.newscientist.com/article/mg20126955.400-north-atlantic-is-worlds-climate-

superpower.html (accessed 24 November 2009)

23 Saved by the Sun New Scientist 16 September 2006 p 32

24 http://en.wikipedia.org/wiki/Solar_cycle (accessed 02 December 2009)

25 Sea Level And Flood Risk Forecasts For Weymouth Harbour Geoff Kirby

(25 November 2009) available for download at

www.geoffkirby.co.uk/CoastalReport.doc or www.geoffkirby.co.uk/CoastalReport.pdf

26 This rise will also be observed throughout Weymouth Bay.


Draft

27 http://www.pol.ac.uk/ntslf/hilo.php?port=weymouth (accessed 02 December 2009)

28 The rainfall at Martinstown was a UK record until November 2009 when it was

exceeded by rain falling in Cumbria.

GJK01305

Keywords:- Weymouth Dorset climate change tides tidal wind rainfall sunspots floods

flooding park district barometric pressure Dorset coast forum

some storm-related weather trends in weymouth · some storm-related weather trends in weymouth...

Documents