A aviation & planes forum. AviationBanter

If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below.

Go Back   Home » AviationBanter forum » rec.aviation newsgroups » General Aviation
Site Map Home Register Authors List Search Today's Posts Mark Forums Read Web Partners

Ice meteors, climate, sceptics



 
 
Thread Tools Display Modes
  #41  
Old February 23rd 04, 02:50 AM
Eric Hocking
external usenet poster
 
Posts: n/a
Default

Brian Sandle wrote in message ...
Eric Hocking wrote:

[...]
I have, in spite of your diversion attempts, tried (and I believe
succeeded) in showing that the crop circle proponent's arguments that
FMD (pedestrian) restrictions had no impact on circle building in 2001
in the UK is unsupportable. What can be concluded from that is up to
those that have been lurking.


I thought river flows might give some indication of weather.


Only of rainfall in the catchment area, surely? While *you* might
think river flow gives some indication of weather, you have not shown
it to be so. The rest of your calculations, while interesting, do not
show anything of the sort.

Why not just check the monthly weather figures from the Government
Meteorological Bureau?

http://www.met-office.gov.uk/climate/uk/

Monthly numbers are available to 1998, but your best bet is from 1999
as the data is tabulated.

http://www.nwl.ac.uk/ih/nrfa/monthly...2/07/rv00.html

gives the flows of a number of UK rivers but unfortunately only from
1999 to 2002. I have tried to estimate the flows from the
logarithmic scales on the diagram for the Itchen river which flows
in Hampshire and might give some indication for the weather
situation in Wiltshire/Hampshire area. If as you say you work with
govt info maybe you know of a better source.


You really do read a lot into other people's posts don't you? I
didn't say I "work with govt info" - other than the references I've
provided in this thread.

And the crop circles I have taken from
http://www.cropcircleresearch.com/ar...tribution.html


Whoa, whoa, whoa. You've jumped from *thinking* that river flow *may*
correlate to weather (uh, what about sunshine and temperature,
especially wrt crops?) straight to *proving* a correlation between
river flow to crop circle emergence and FMD? Try showing the logic of
this before attempting to force the numbers.

Year 1999 2000 2001 2002

Mar+Ap+May flow 18 24 35 22
Apr+May circs 24 14 9 4

Mar+Ap flow 13 15 25 15
Apr circs 9 3 0 1

FMD Yes(1)/No(0)0 0 1 0
"
This amount of data is not really sufficient, but it is interesting
what turns up is a -0.49 correlation between Itchen river Mar Apr
May flows and Apr+May crop circles {call it r(flows-circles)}. I am
risking using the Pearson correlation. And the Mar+Apr flows and the
Apr circles correlation is -0.67.


You're risking more than choosing the correct correlation technique.
You've yet to show that the Itchen River flow has any relationship to
weather conditions. As for river flow, have you factored in seasonal
abstraction from river systems?

Also there turns up a correlation of river flow to FMD

Jan Feb Mar Apr May
0.94 0.96 0.97 0.97 0.68

May being when it was finished there?


This "correlation" implies what? River flow affects government
decisions on lifting FMD restrictions?

But anyway taking the Mar Apr May flows figures, since weather might
stop hoaxers,


You are yet to show that the Itchen River flow readings actually has
any reflection on the weather pattern trends.

r(flows-fmd) = 0.94.
And is there a correlation between FMD & circles?
Yes, r(fmd-circles) = -0.29, a small negative correlation, rather
less than from above
r(flows-circles) = -0.49.

Then what happens when partial correlation is used to get a feel for
removing affects of the factors?

When the effects of the rivers are nullified then FMD becomes
*positively* related to circles.

r(fmd-circles.flows) = 0.57 instead of -0.29
and for completeness
r(flows-fmd.circles) = 0.96 instead of 0.94, no change, rather
indicating circles not causative,
r(flows-circles.fmd) = -0.66 instead of -0.49, not much
change indicating FMD not really causative.
With that small amount of data, so far, some of that could be by
chance.


And also shows that you can "prove" anything with forced numbers and
illogical connections. You need first to show that there is a logical
connection between a single river's flow trend as an indicator of
weather conditions (rainfall, temperature, sunshine).

How about using *weather* data directly for the county, instead of
attempting to derive this data from a single river monitoring station?

--
Eric Hocking
  #42  
Old February 23rd 04, 04:33 AM
Brian Sandle
external usenet poster
 
Posts: n/a
Default

Eric Hocking wrote:
Brian Sandle wrote in message ...
Eric Hocking wrote:

[...]
I have, in spite of your diversion attempts, tried (and I believe
succeeded) in showing that the crop circle proponent's arguments that
FMD (pedestrian) restrictions had no impact on circle building in 2001
in the UK is unsupportable. What can be concluded from that is up to
those that have been lurking.


I thought river flows might give some indication of weather.


Only of rainfall in the catchment area, surely? While *you* might
think river flow gives some indication of weather, you have not shown
it to be so.


They may be a better indication than weather of what the ground is like. They
do not increase flow until the ground is saturated. (Though, not appropriate
to UK in my knowledge, fast run off can occur off baked land.)

The rest of your calculations, while interesting, do not
show anything of the sort.


Why not just check the monthly weather figures from the Government
Meteorological Bureau?


http://www.met-office.gov.uk/climate/uk/


Monthly numbers are available to 1998, but your best bet is from 1999
as the data is tabulated.


It would be complex factoring in evapotranspiration. Since there are only 4
years it is not really worth it.

http://www.nwl.ac.uk/ih/nrfa/monthly...2/07/rv00.html

gives the flows of a number of UK rivers but unfortunately only from
1999 to 2002. I have tried to estimate the flows from the
logarithmic scales on the diagram for the Itchen river which flows
in Hampshire and might give some indication for the weather
situation in Wiltshire/Hampshire area. If as you say you work with
govt info maybe you know of a better source.


You really do read a lot into other people's posts don't you? I
didn't say I "work with govt info" - other than the references I've
provided in this thread.


And the crop circles I have taken from
http://www.cropcircleresearch.com/ar...tribution.html


Whoa, whoa, whoa. You've jumped from *thinking* that river flow *may*
correlate to weather (uh, what about sunshine and temperature,
especially wrt crops?)


Extra flows indicate the ground cannot hold the water, therefore there has
been less sunshine and temperature.

straight to *proving* a correlation between
river flow to crop circle emergence and FMD? Try showing the logic of
this before attempting to force the numbers.


How am I `forcing' the numbers?

Year 1999 2000 2001 2002

Mar+Ap+May flow 18 24 35 22
Apr+May circs 24 14 9 4

Mar+Ap flow 13 15 25 15
Apr circs 9 3 0 1

FMD Yes(1)/No(0)0 0 1 0
"
This amount of data is not really sufficient, but it is interesting
what turns up is a -0.49 correlation between Itchen river Mar Apr
May flows and Apr+May crop circles {call it r(flows-circles)}. I am
risking using the Pearson correlation. And the Mar+Apr flows and the
Apr circles correlation is -0.67.


You're risking more than choosing the correct correlation technique.
You've yet to show that the Itchen River flow has any relationship to
weather conditions. As for river flow, have you factored in seasonal
abstraction from river systems?


It is only very rough. Besides seasonal effects should be similar from year
to year and factor out.


Also there turns up a correlation of river flow to FMD

Jan Feb Mar Apr May
0.94 0.96 0.97 0.97 0.68

May being when it was finished there?


This "correlation" implies what?



That for some reason Itchen river flow was high at the same time FMD was
present.

River flow affects government
decisions on lifting FMD restrictions?


Presumably the restricitions were lifted when it was thought there was less
risk. I doubt there would have been any talk of rivers transporting FMD. When
the land dried a bit stock could get out into the fields and have a bit less
close contact and so less chance for transmission of FMD.


But anyway taking the Mar Apr May flows figures, since weather might
stop hoaxers,


You are yet to show that the Itchen River flow readings actually has
any reflection on the weather pattern trends.


Next you will be asking me to prove that day is going to be lighter than
night.

r(flows-fmd) = 0.94.
And is there a correlation between FMD & circles?
Yes, r(fmd-circles) = -0.29, a small negative correlation, rather
less than from above
r(flows-circles) = -0.49.

Then what happens when partial correlation is used to get a feel for
removing affects of the factors?

When the effects of the rivers are nullified then FMD becomes
*positively* related to circles.

r(fmd-circles.flows) = 0.57 instead of -0.29


which indicates flows are connected to cause.

and for completeness
r(flows-fmd.circles) = 0.96 instead of 0.94, no change, rather
indicating circles not causative,
r(flows-circles.fmd) = -0.66 instead of -0.49, not much
change indicating FMD not really causative.
With that small amount of data, so far, some of that could be by
chance.


And also shows that you can "prove" anything with forced numbers and
illogical connections.



Here is the formula for you to have some fun:

r(po.y)= [r(po)-r(py).r(oy)]/sqr.root[1-{r(py)}^2].sqr.root[1-{r(oy)}^2]

(Bruning & Kintz).

where r(po.y) is the partial correlation between p and o, partialling out y.

r(po) is the non-partial correlation between p and o, &c for p & y, o & y.


When the partial correlation tends to zero that means the partialled out
variable is causal and the non-partial correlation is spurious.

When the partial correlation is no different from the non-partial, that means
the partailled out variable is not causal.


You need first to show that there is a logical
connection between a single river's flow trend as an indicator of
weather conditions (rainfall, temperature, sunshine).


How about using *weather* data directly for the county, instead of
attempting to derive this data from a single river monitoring station?


The first crop circle for 2001 was in Hampshire at latitude 50 deg 58.6 min
north, longitude 1 deg 5.9 mins west. That is only 10 or 20 miles from the
Itchen river (which has its mouth near Southampton). It is not a big reiver
and seems to have its source on the same side of South Downs.
  #43  
Old February 23rd 04, 10:36 PM
Eric Hocking
external usenet poster
 
Posts: n/a
Default

Brian Sandle wrote in message ...
Eric Hocking wrote:
Brian Sandle wrote in message ...
Eric Hocking wrote:
[...]
I have, in spite of your diversion attempts, tried (and I believe
succeeded) in showing that the crop circle proponent's arguments that
FMD (pedestrian) restrictions had no impact on circle building in 2001
in the UK is unsupportable. What can be concluded from that is up to
those that have been lurking.

I thought river flows might give some indication of weather.


Only of rainfall in the catchment area, surely? While *you* might
think river flow gives some indication of weather, you have not shown
it to be so.


They may be a better indication than weather of what the ground is like.


You're saying that a single river's flow pattern taken from one
station on the river is a better indicator of a region's weather that
the weather data gathered by the Bureau of Meteorological's database
from over 30 stations throughout the region?
http://www.met-office.gov.uk/climate/uk/networks/climnet2.html

That's not logical.

Why is it that you chose to ignore the rainfall data at the site you
gleaned the river flow data from, as well as the hydrological summary
put together by experts AND the groundwater discussion.

Why is that Brian? If I were a cynical person I might mention the
phrase "data mining". Why don't you put your (incorrect) assumption
on the data into the full context of the information provided.

They
do not increase flow until the ground is saturated. (Though, not appropriate
to UK in my knowledge, fast run off can occur off baked land.)


Your knowledge of UK weather and water reserves is limited then - and
you would have realised this if you had actually read the monthly
summaries relating to rainfall, river flow and groundwater on the site
you decided to ONLY use river flow data from.

FYI. Earlier this year we experienced torrential rain after a longer
and drier than usual summer in 2003. Water companies reported that
the rivers/reservoirs and especially groundwater reserves were not
being replenished because of the rain running of the dry earth.
Again, your assumptions do not reflect the facts.

The rest of your calculations, while interesting, do not
show anything of the sort.
Why not just check the monthly weather figures from the Government
Meteorological Bureau?
http://www.met-office.gov.uk/climate/uk/
Monthly numbers are available to 1998, but your best bet is from 1999
as the data is tabulated.


It would be complex factoring in evapotranspiration.


Sod complex factoring - all you want to know is weather conditions in
the region. This is supplied in better detail and accuracy by weather
station data than unsound derivations from river flow data that you do
not completely understand.

Since there are only 4
years it is not really worth it.


You obviously didn't look or try too hard. The data sets go back to
1855, and the three prime weather indicators (as far as crops may be
concerned, rainfall, hours of sunshine and min/max temperatures) are
available back to 1895.
ie http://www.met-office.gov.uk/climate/uk/stationdata/southamptondata.txt

snip
Whoa, whoa, whoa. You've jumped from *thinking* that river flow *may*
correlate to weather (uh, what about sunshine and temperature,
especially wrt crops?)


Extra flows indicate the ground cannot hold the water,


Incorrect and not just for physical reasons stated above.
There are economical, political and social conditions that can affect
river flow. Not the least of which is the changes in water
abstraction licensing in the UK, let alone changes in irrigation
practices in the UK agricultural industry.

Your entire set of calculations are based on incorrect assumptions and
therefore of little use.

been less sunshine and temperature.


As shown above, if you do not fully understand the data gathering
process, let alone the data itself, adding in external factors such as
pressures on water resources in a region, why attempt to derive
weather conditions, probably incorrectly, when you have a WEATHER
resource available to you?

Here's a timeline of weather conditions and crops in Hampshire.
/

I'll do the sums for you:
Month Sun Mean Cum.
Jan Hours Temp Rainfall
-Aug (C) (mm)

1999 1102 10.7 771
2000 1127 10.4 732
2001 1193 10.1 719
2002 1054 10.6 854

Well whaddayano? 2001 had the MOST sunshine, the LOWEST mean
temperature and LEAST rainfall for those years. I'll fully admit that
this is for the entire SW, but it does not gel with your conclusions
very well at all. Oh, before you ask, numbers for Mar, Apr ,May a

1999 406 9.5 259
2000 451 8.8 306
2001 454 8.5 288
2002 466 9.1 300

straight to *proving* a correlation between
river flow to crop circle emergence and FMD? Try showing the logic of
this before attempting to force the numbers.


How am I `forcing' the numbers?


See below, but first, show the logic of your reasoning that river flow
*should* have any relationship to crop circle timing in any year.

Year 1999 2000 2001 2002
Mar+Ap+May flow 18 24 35 22
Apr+May circs 24 14 9 4


For one thing, you make the same mistake as on the researcher's page.
You cannot use the total number of circles recorded in a month to show
a timeline trend. Using an example I've used befo

I have 4 fields and one month (say May) a circle is built in each one
on the 1st.
By your method 4 circles are recorded for May.

Next year, 3 of the fields are closed to the public, but 4 circles are
built in the remaining field, on the 31st.
By your method 4 circles are recorded for May.

This is further compounded by you combining the totals, so yes, you
are forcing the numbers, or at least misrepresenting them for
correlation you are attempting to show.

The issue is, and has always been, *timing* of the appearance not
number of appearances.

Mar+Ap flow 13 15 25 15
Apr circs 9 3 0 1

FMD Yes(1)/No(0)0 0 1 0


Incorrect data there. If FMD Yes/No is supposed to indicate where FMD
restrictions are in place, you'd be better of showing month by month
trends. Total for the year is ridiculous, it shows nothing.

snip
You're risking more than choosing the correct correlation technique.
You've yet to show that the Itchen River flow has any relationship to
weather conditions. As for river flow, have you factored in seasonal
abstraction from river systems?


It is only very rough. Besides seasonal effects should be similar from year
to year and factor out.


The whole point of this part of the thread was for you to show that
changes in weather conditions from year-to-year affect the timing of
the appearance of circles in the UK. You are now saying that the
weather, being similar from year to year, factors OUT of the
equations?

Which is it Brian. Does weather affect the appearance of circles or
not. If not - what was all the statistics rubbish in aid of?


Also there turns up a correlation of river flow to FMD


You are yet to provide a logical and reasonable justification for
attempting to *find* a correlation. Until you do this is just
numerology.

Jan Feb Mar Apr May
0.94 0.96 0.97 0.97 0.68


Different set of data again. You go on later to compare annual totals
with monthly correlations yet you do not show the working.

May being when it was finished there?

This "correlation" implies what?


That for some reason Itchen river flow was high at the same time FMD was
present.


One last one - why did you only use the Itchen River data when there
is more than over a dozen stations in the Hampsire area you could have
included? Not only that, you take the data in isolation. Try putting
it in context with the rest of the available hydrological data
available at that site.

River flow affects government
decisions on lifting FMD restrictions?


Presumably the restricitions were lifted when it was thought there was less
risk. I doubt there would have been any talk of rivers transporting FMD. When
the land dried a bit stock could get out into the fields and have a bit less
close contact and so less chance for transmission of FMD.


Show a cite for this presumption. You won't be able to, because it is
wrong, but I'd like to see how you came to this incorrect assumption.
I documented exactly what the determination process by the government
was. THis is pure obfuscation, Brian.

But anyway taking the Mar Apr May flows figures, since weather might
stop hoaxers,


You are yet to show that the Itchen River flow readings actually has
any reflection on the weather pattern trends.


Next you will be asking me to prove that day is going to be lighter than
night.


At least that would be logic that any reasonable person could follow.
But since YOU are putting forward the contention that the river
station data CAN be used to reflect weather pattern trends, it IS up
to you to show it.

You went up this particular creek - and I think you'll find you forgot
the paddle.

r(flows-fmd) = 0.94.
And is there a correlation between FMD & circles?
Yes, r(fmd-circles) = -0.29, a small negative correlation, rather
less than from above
r(flows-circles) = -0.49.
Then what happens when partial correlation is used to get a feel for
removing affects of the factors?
When the effects of the rivers are nullified then FMD becomes
*positively* related to circles.

r(fmd-circles.flows) = 0.57 instead of -0.29


which indicates flows are connected to cause.


No it doesn't - it's just illogical numerology (how's THAT for an
tautology?).
It's also a contradiction of your own conviction that correlation
isn't causation.
Again I ask, which is it Brian?

and for completeness
r(flows-fmd.circles) = 0.96 instead of 0.94, no change, rather
indicating circles not causative,
r(flows-circles.fmd) = -0.66 instead of -0.49, not much
change indicating FMD not really causative.
With that small amount of data, so far, some of that could be by
chance.

And also shows that you can "prove" anything with forced numbers and
illogical connections.

Here is the formula for you to have some fun:

snip

You need to show that there is a logical connection between the data
befor you can attempt to derive and analyse correlations. You've yet
to do so.

Manipulating the data in isolation and also ignoring the other station
data doesn't help either.

When the partial correlation tends to zero that means the partialled out
variable is causal and the non-partial correlation is spurious.
When the partial correlation is no different from the non-partial, that means
the partailled out variable is not causal.


When the data is taken in isolation, misunderstood and incomplete, ALL
conclusions are spurious.

You need first to show that there is a logical
connection between a single river's flow trend as an indicator of
weather conditions (rainfall, temperature, sunshine).


How about using *weather* data directly for the county, instead of
attempting to derive this data from a single river monitoring station?


The first crop circle for 2001 was in Hampshire at latitude 50 deg 58.6 min
north, longitude 1 deg 5.9 mins west. That is only 10 or 20 miles from the
Itchen river (which has its mouth near Southampton). It is not a big reiver
and seems to have its source on the same side of South Downs.


.... and you ignore the dozen or so other station data.
.... and you are yet to show a logical connection between river flow
and weather
.... and you are yet to show a logical connection between weather and
FMD
.... and you are yet to show a logical connection between river flow
and FMD

I suggest you read the guidelines and summaries on the NRFA site
before you waste any more time crunching unconnected data.

--
Eric Hocking
  #44  
Old February 24th 04, 01:27 AM
Brian Sandle
external usenet poster
 
Posts: n/a
Default

In alt.freemasonry Eric Hocking wrote:
Brian Sandle wrote in message ...
Eric Hocking wrote:
Brian Sandle wrote in message ...
Eric Hocking wrote:
[...]
I have, in spite of your diversion attempts, tried (and I believe
succeeded) in showing that the crop circle proponent's arguments that
FMD (pedestrian) restrictions had no impact on circle building in 2001
in the UK is unsupportable. What can be concluded from that is up to
those that have been lurking.

I thought river flows might give some indication of weather.


Only of rainfall in the catchment area, surely? While *you* might
think river flow gives some indication of weather, you have not shown
it to be so.


They may be a better indication than weather of what the ground is like.


You're saying that a single river's flow pattern taken from one
station on the river


The river sums the water as it comes in. If any is being taken out I presume
it will be less in wet weather.

is a better indicator of a region's weather that
the weather data gathered by the Bureau of Meteorological's database
from over 30 stations throughout the region?
http://www.met-office.gov.uk/climate/uk/networks/climnet2.html


The Itchen river is closer to the Hamphsire site than most of those stations.
There are not 30 in the region - only a few, and mainly near the coast.


That's not logical.


Why is it that you chose to ignore the rainfall data at the site you
gleaned the river flow data from, as well as the hydrological summary
put together by experts AND the groundwater discussion.


I was just wanting a quick result.

Ground water is a bit deeper. I think it takes longer to refill. That
happens when stress is taken off the wells. Rainfall, as I said, sums into
the river.

Why is that Brian? If I were a cynical person I might mention the
phrase "data mining".


I was looking for the weather and came across the river data. I did not go
searching out what fits to that point. I did then take a look at which months
of river flow fitted best. More on data choice below.

Why don't you put your (incorrect) assumption
on the data into the full context of the information provided.


I think it is pretty good.

They
do not increase flow until the ground is saturated. (Though, not appropriate
to UK in my knowledge, fast run off can occur off baked land.)


Your knowledge of UK weather and water reserves is limited then - and
you would have realised this if you had actually read the monthly
summaries relating to rainfall, river flow and groundwater on the site
you decided to ONLY use river flow data from.


Comment on your data below.

FYI. Earlier this year we experienced torrential rain after a longer
and drier than usual summer in 2003. Water companies reported that
the rivers/reservoirs and especially groundwater reserves were not
being replenished because of the rain running of the dry earth.


Yes the water was being lost. The reservoirs would still have to be used.

Again, your assumptions do not reflect the facts.


I put that bit in brackets because it happens later in the season. I didn't
realise an island country with a latitude like UK's would get very dry.

The rest of your calculations, while interesting, do not
show anything of the sort.
Why not just check the monthly weather figures from the Government
Meteorological Bureau?
http://www.met-office.gov.uk/climate/uk/
Monthly numbers are available to 1998, but your best bet is from 1999
as the data is tabulated.


It would be complex factoring in evapotranspiration.


Sod complex factoring - all you want to know is weather conditions in
the region.


The weather was introduced as an explanation for late crops. But river flows
sum the weather and give even more of a picture than the weather.

This is supplied in better detail and accuracy by weather
station data than unsound derivations from river flow data that you do
not completely understand.


But there are few in the area.

Since there are only 4
years it is not really worth it.


You obviously didn't look or try too hard. The data sets go back to
1855, and the three prime weather indicators (as far as crops may be
concerned, rainfall, hours of sunshine and min/max temperatures) are
available back to 1895.
ie http://www.met-office.gov.uk/climate/uk/stationdata/southamptondata.txt


I might look for river flows back further and relate crops circles.

snip
Whoa, whoa, whoa. You've jumped from *thinking* that river flow *may*
correlate to weather (uh, what about sunshine and temperature,
especially wrt crops?)


Extra flows indicate the ground cannot hold the water,


Incorrect and not just for physical reasons stated above.
There are economical, political and social conditions that can affect
river flow. Not the least of which is the changes in water
abstraction licensing in the UK, let alone changes in irrigation
practices in the UK agricultural industry.


More will be taken out in dry weather.

Your entire set of calculations are based on incorrect assumptions and
therefore of little use.


Where is a measure of surface ground saturation? And I not fungi do not grow
until the ground dries a bit. Not that I am saying it is fungi, but that in
some cases it may be.

been less sunshine and temperature.


As shown above, if you do not fully understand the data gathering
process, let alone the data itself, adding in external factors such as
pressures on water resources in a region, why attempt to derive
weather conditions, probably incorrectly, when you have a WEATHER
resource available to you?


Here's a timeline of weather conditions and crops in Hampshire.
/


I'll do the sums for you:
Month Sun Mean Cum.
Jan Hours Temp Rainfall
-Aug (C) (mm)


1999 1102 10.7 771
2000 1127 10.4 732
2001 1193 10.1 719
2002 1054 10.6 854


Well whaddayano? 2001 had the MOST sunshine, the LOWEST mean
temperature and LEAST rainfall for those years. I'll fully admit that
this is for the entire SW, but it does not gel with your conclusions
very well at all. Oh, before you ask, numbers for Mar, Apr ,May a


Yes, why did you give up to August? I did not look at river flows that far.

1999 406 9.5 259
2000 451 8.8 306
2001 454 8.5 288
2002 466 9.1 300


From your graph, thanks, the rain eased by May in 2001. That would drop the
total for those three months.

The slope of the cumulative rainfall graph shows the amount of rain effect.
Note a long constant slope in 2001, starting from only 100 in mid Jan and
going to nearly 500 by mid Apr. And only about 50 hours of sun to mid March.
There was not much sun in 2002 either in that time, but the slope of the
rainfall graph tapered off. Circles were also a bit few in 2002, BTW.


straight to *proving* a correlation between
river flow to crop circle emergence and FMD? Try showing the logic of
this before attempting to force the numbers.


How am I `forcing' the numbers?


See below, but first, show the logic of your reasoning that river flow
*should* have any relationship to crop circle timing in any year.


I need some help with statistical education.

There is always the warning against doing lots of correlations and taking the
ones you think you like. If there is 5% chance of any particular result
happening by chance then the more you take the more likely your result is by
chance.

So you are told to think out a theory and test it by stats.

I feel that that is now rather hit and miss. Your theory and the stats still
have a 5% chance of being wrong. What is needed is an analysis which deals
with that. That usually means wait until 20 more people have repeated your
work, doesn't it?

My approach would be to look for all correlations with knowledge that 5% are
by chance, to acknowledge that but to try to build a pattern of what is
happening using the partial correlations to help, and to decide on further
directions, not to wait the years for people to check the one correlation you
have done.

I tend to develop reasoning after seeing a pattern.

Year 1999 2000 2001 2002
Mar+Ap+May flow 18 24 35 22
Apr+May circs 24 14 9 4


For one thing, you make the same mistake as on the researcher's page.
You cannot use the total number of circles recorded in a month to show
a timeline trend.


If there are fewer circles then they are likely to start later.

Using an example I've used befo


I have 4 fields and one month (say May) a circle is built in each one
on the 1st.
By your method 4 circles are recorded for May.


Next year, 3 of the fields are closed to the public, but 4 circles are
built in the remaining field, on the 31st.
By your method 4 circles are recorded for May.


Yes, but fewer show Apr - May.

This is further compounded by you combining the totals, so yes, you
are forcing the numbers, or at least misrepresenting them for
correlation you are attempting to show.


The issue is, and has always been, *timing* of the appearance not
number of appearances.


I could go back and give 0 for none that month and 1 if any occurred. Might
not be too different.

Mar+Ap flow 13 15 25 15
Apr circs 9 3 0 1

FMD Yes(1)/No(0)0 0 1 0


Incorrect data there. If FMD Yes/No is supposed to indicate where FMD
restrictions are in place, you'd be better of showing month by month
trends. Total for the year is ridiculous, it shows nothing.


It is not total for year it is only if FMD was present in the tiem just
preceding when the circles normally appear.

snip
You're risking more than choosing the correct correlation technique.
You've yet to show that the Itchen River flow has any relationship to
weather conditions. As for river flow, have you factored in seasonal
abstraction from river systems?


It is only very rough. Besides seasonal effects should be similar from year
to year and factor out.


The whole point of this part of the thread was for you to show that
changes in weather conditions from year-to-year affect the timing of
the appearance of circles in the UK. You are now saying that the
weather, being similar from year to year, factors OUT of the
equations?


Which is it Brian. Does weather affect the appearance of circles or
not. If not - what was all the statistics rubbish in aid of?


Seasonal abstraction I thought meant that more water is drawn in the dry
season of each year. In a wet year that will be below average.

Also there turns up a correlation of river flow to FMD


You are yet to provide a logical and reasonable justification for
attempting to *find* a correlation. Until you do this is just
numerology.


See above theory.


Jan Feb Mar Apr May
0.94 0.96 0.97 0.97 0.68


Different set of data again. You go on later to compare annual totals
with monthly correlations yet you do not show the working.


Just giving the individual months before combining.

May being when it was finished there?
This "correlation" implies what?


That for some reason Itchen river flow was high at the same time FMD was
present.


One last one - why did you only use the Itchen River data when there
is more than over a dozen stations in the Hampsire area you could have
included?


Not in your map on that side of South Downs.

Not only that, you take the data in isolation. Try putting
it in context with the rest of the available hydrological data
available at that site.


Where is the surface ground saturation data?

River flow affects government
decisions on lifting FMD restrictions?


Presumably the restricitions were lifted when it was thought there was less
risk. I doubt there would have been any talk of rivers transporting FMD. When
the land dried a bit stock could get out into the fields and have a bit less
close contact and so less chance for transmission of FMD.


Show a cite for this presumption. You won't be able to, because it is
wrong, but I'd like to see how you came to this incorrect assumption.


They have less contact with the other animals' excrement or drinking water so
probability of rapid spread reduces.

I documented exactly what the determination process by the government
was. THis is pure obfuscation, Brian.


I just presumed that when the disease was controlled the restricitons were
removed. It did not become controlled while the animals were cramped
together.


But anyway taking the Mar Apr May flows figures, since weather might
stop hoaxers,


You are yet to show that the Itchen River flow readings actually has
any reflection on the weather pattern trends.


Next you will be asking me to prove that day is going to be lighter than
night.


At least that would be logic that any reasonable person could follow.
But since YOU are putting forward the contention that the river
station data CAN be used to reflect weather pattern trends, it IS up
to you to show it.


The important summed weather effect is likely to be reflected in river flows.

You went up this particular creek - and I think you'll find you forgot
the paddle.


Allow me to associate ideas. A paddle control on an old computer could be
hooked up to record surface ground resistance therefore water saturation.

r(flows-fmd) = 0.94.
And is there a correlation between FMD & circles?
Yes, r(fmd-circles) = -0.29, a small negative correlation, rather
less than from above
r(flows-circles) = -0.49.
Then what happens when partial correlation is used to get a feel for
removing affects of the factors?
When the effects of the rivers are nullified then FMD becomes
*positively* related to circles.

r(fmd-circles.flows) = 0.57 instead of -0.29


which indicates flows are connected to cause.


No it doesn't - it's just illogical numerology (how's THAT for an
tautology?).


I didn't say they are cause. Somehow they are likely to be connected to it.

It's also a contradiction of your own conviction that correlation
isn't causation.
Again I ask, which is it Brian?


Correlation is not causation, but partial correlation hints at causation.

and for completeness
r(flows-fmd.circles) = 0.96 instead of 0.94, no change, rather
indicating circles not causative,
r(flows-circles.fmd) = -0.66 instead of -0.49, not much
change indicating FMD not really causative.
With that small amount of data, so far, some of that could be by
chance.
And also shows that you can "prove" anything with forced numbers and
illogical connections.

Here is the formula for you to have some fun:

snip


Snipped the formula? Don't you like it?

You need to show that there is a logical connection between the data
befor you can attempt to derive and analyse correlations. You've yet
to do so.


See again, above theory.

Manipulating the data in isolation and also ignoring the other station
data doesn't help either.


There wasn't any closer. Also you included irrelevant months.

When the partial correlation tends to zero that means the partialled out
variable is causal and the non-partial correlation is spurious.
When the partial correlation is no different from the non-partial, that means
the partailled out variable is not causal.


When the data is taken in isolation, misunderstood and incomplete, ALL
conclusions are spurious.


That is not the technical sense of the word. As I said the old way is to wait
for people to repeat your experiment. I think there is better way to do
things.

You need first to show that there is a logical
connection between a single river's flow trend as an indicator of
weather conditions (rainfall, temperature, sunshine).


How about using *weather* data directly for the county, instead of
attempting to derive this data from a single river monitoring station?


The first crop circle for 2001 was in Hampshire at latitude 50 deg 58.6 min
north, longitude 1 deg 5.9 mins west. That is only 10 or 20 miles from the
Itchen river (which has its mouth near Southampton). It is not a big reiver
and seems to have its source on the same side of South Downs.


... and you ignore the dozen or so other station data.


list them.

... and you are yet to show a logical connection between river flow
and weather


It shows on your diagram.

... and you are yet to show a logical connection between weather and
FMD


Many diseases are seasonal. Humans can get more fungal diseases in hot humid
weather when the sweat provides the environment.

Lack of sunlight reduces vitamin D an important immune regulator, helpful
against TB I think.

... and you are yet to show a logical connection between river flow
and FMD


Cooped up animals meant it could be spread easily and rapidly.

I suggest you read the guidelines and summaries on the NRFA site
before you waste any more time crunching unconnected data.


National Rural Fire Authority?

Now if I were dealing with a summer disease I might correlate to fire risk,
since that would indicate moisture.
 




Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT +1. The time now is 09:26 AM.


Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Copyright ©2004-2024 AviationBanter.
The comments are property of their posters.