NavList:

   

Reply

> Thanks! You mentioned a serious problem indeed.
> The local joke is
> "If you don't like the weather in Kiel, just wait 10 minutes".
>
> And this is literally true.
> I've never seen a place where clouds come and go so fast.
> In the limited time that Sun is visible from my window,
> it is very hard to catch it in a hole between the clouds.

Alex

Very much like my experience on Lake Michigan last weekend.  The marine
forecasts should have been done for every ten-mile stretch, and even then
would have changed almost hourly.  Clouds sweeping in from the southwest.
The plot of hourly barometric pressure looked like a sine wave, and the
horizon was hazy enough that it was nearly impossible to get a good IE check
(air and water temp maybe 20 F apart).  Out of five tries, two IE checks
were in the right neighborhood at 0.6' on the arc (usual O.7'), one was 0.2'
off the arc, one at 1.0' on the arc, and the last 1.2' on the arc.  Tried
every trick I have read, including sweeping the horizon once set and looking
for a traveling "hump," all to no avail.  Winds had started about 17 kn out
of Michigan City Saturday, dropped to 7 kn an hour or two later, and then
picked up to 32 kn with heavy rain and lightning about three hours into the
sail to Chicago (about 30 minutes after my observations). Off and on.
(Talked to some of the racers in the Chicago/Michigan City race after the
fact and they saw breezes in the lower 40 kn range--as forecast in the storm
warnings. Tornadoes were reported in Michigan.)  Quite an interesting day
weather wise.

Despite my disappointment at the lack of a crisp horizon, I was not going to
lose the opportunity to shoot on the water, and reckoned I was not the first
sailor to encounter the horizon obstacle. (Remembered tales of  tropical
navigators with clear skies but a hazy horizon, but did not have a dish of
mercury with me ;-) Seas were calm with 1-3 ft swells, so I had a go at it.
Picked the closest to a sharp line I could find and used it.

Besides the horizon, yaw was the biggest obstacle to overcome.  Adjusting
for pitch and heel/roll was relatively easy, but yaw with an inexperienced
guest helmsman was difficult. Sort of "set and wait, adjust, set and wait."

Before departing I took measurements from the waterline to the cockpit sole,
coachroof by the mast, and foredeck at the anchor locker for use in
calculating dip.  Then it was a simple matter of measuring my height of eye
from the surface, and adjusting for heel by noting the difference between
resting waterline and boot-topping/water under sail.  I had a mate hold my
GPS and mark down the time and coordinates at my mark.  Then I averaged the
times, Ho's, and position and did a single reduction with the averages on a
pocket calculator.  Results were 1.9' Away.

Monday on the return trip the same "spot weather" but milder.  When we left
for Michigan City small-craft warnings were forecast there for the
afternoon.  We actually wound up motor sailing in due to lack of wind. It
was hot and humid (and when the wind died we got a boatload of those nasty
little biting flies that can't read the insect-repellant label that states
it repels them--I prefer the Canadian Wren-sized flies you can hunt with a
410 shotgun/birdshot).  2-3 foot swells, wind still out of the southwest.
The horizon was like a line of loose cotton between the water and sky with
no discernable line, except to the northwest where it resembled Saturday's
horizon.  We were about 9 nm off, and with the haze against the sand dunes
it seemed like a lost cause. I tried one shot, and decided to give it up.
Then I thought there might be something to be learned about accuracy and
repeatability under the conditions, so did three shots and recorded them.
The three averaged out to 0.8' Away.  Better lucky than good.

When I returned home I ran all the individual observations through Omar
Reis's site (http://www.tecepe.com.br/scripts/AlmanacPagesISAPI.isa) for
Hc's , calculated individual errors and averaged them, and used Excel to
plot the Hc's vs. Ho's, and established a linear-regression trend line for
the Ho's.

Saturday, June 4, approx. 13:36:48 EST
On-board calculation of averages: 1.9' Away

Error by Omar (Ho - Hc)
-1.1'
-1.8'
-2.4'
-3.2
-0.3
Mean of errors:  -1.7 (Away)

The slope of the Excel-generated linear-regression trend line for Ho was
parallel to the Hc slope plot but below.  I would love to discard the -3.2
as an outlier and improve the results significantly, but in all fairness
would also have to discard the -0.3 when compared to the Hc slope if
plotting by hand on a boat, so little would change.

Sunday, June 6, approx. 14:33:45 EST
On-board calculation of averages: 0.8 Away

Error by Omar (Ho - Hc)
-6.3'
-5.0'
+8.7'
Mean -0.9 (Away)

As you can imagine, this Excel plot was a mess.  The slope of the Hc line
was much steeper than the slope of the Ho trend line.  By chance I guess
(being mathematically challenged), the lines intersected each other in their
middles, so luck won out?  Fitting to the Hc slope would have given similar
results.  Somehow, founded or not, my gut tells me I am better to fit the
data points to the Hc slope then picking an arbitrary time/elevation than
doing the same on the linear-regression slope.

The difference in Hc calculations from my averages and Omar's site Hc's I
regard as rounding errors.  Often my calculations from the Nautical Almanac
data vs. Omar's are 0.1' off.

Also played with position with my hockey-puck compass and sextant using the
Sears Tower and John Hancock buildings around Chicago, and position from an
object (Sears Tower and Michigan City power-plant smokestack using the
sextant and trig with mixed results.  All in all, a great learning
experience.

Looking forward to your comments and results with a natural horizon and/or
small craft.  Did you pack that magnificent Soviet three-armed protractor to
play with?

Bill

   

Reply