shmoo
Well-Known Member
Its the time of year to plan passages. Most years we try to get away from the East Coast to somewhere we can see through the water. We only get a month and if we are going to get along the south coast and across to Brittainy we need to plan carefully. After pouring over the tidal atlas for hours the inevitable questions like
"What if we left Dover 3 hours later?" or
"If we left on the Thursday, what day would we arrive?"
"No, I meant Friday...."
"Can we get there for Bastille day?"
send me to the edge of sanity (not far distant in my case).
Finally I got round to writing a program to do it. It naively does the old Yachtmaster thing:-
- For each leg of the trip, divide it into hours (at some boat speed),
- for each hour,
-- find the nearest tidal diamond,
-- get time of high tide at the diamond's standard port,
-- find the "tidal hour",
-- interpolate the rate,
-- find the resultant vector
- Add all the vectors to give a CTS.
- Do the whole thing again to get hourly EPs to make sure the whole leg is conducted in the water.
Ideally now you do do it again for each of the next twelve hours to find the shortest passage time. Try it for slightly different boat speeds...
Conceptually simple, but very time consuming and error-prone. A PC will march through the spherical-surface trig sums at millions a sec.
Here is were the program comes in.
Tidal data for standard ports is imported (by the user) from WXtide32 and users enter the tidal diamond data from their charts: the more diamonds you add the better the estimates. For cross channel one at each end and one in the middle seems to work. Coasting along the south coast one at each headland and one in between each seems to work well. A "library" of diamonds is built up over time.
"Routes" (of "legs") can be imported from Seaclear or typed in and the
resulting hourly EPs can be exported to Seaclear (as a "track").
Here is a route from Dover to Newhaven (imported from Seaclear and some of the default WP names have been left unedited)
and the whatif (next 12 hours)
Clearly the time to leave is around 16-17h
Here is the leg detail showing estimated duration of each leg and CTS. Also show is the optional expansion to show which hours of which diamonds the leg estimates are based on.
Finally you can import the hourly EPs back into Seaclear to display them. This is a different example, which shows the EP track more clearly than when coasting.
Its very early days yet. I needed work up a "real" program in C# in a Microsoft-free development environment, for reasons unrelated to sailing, and I knew I would do it better if it was interesting.
There are systems which do this but they are expensive and depend on getting new "packs" for new cruising areas. This scheme depends on data entered by the user, from charts.
Questions for the panel are:
Is this sensible?
Would folks pay modest sums of money for it?
"What if we left Dover 3 hours later?" or
"If we left on the Thursday, what day would we arrive?"
"No, I meant Friday...."
"Can we get there for Bastille day?"
send me to the edge of sanity (not far distant in my case).
Finally I got round to writing a program to do it. It naively does the old Yachtmaster thing:-
- For each leg of the trip, divide it into hours (at some boat speed),
- for each hour,
-- find the nearest tidal diamond,
-- get time of high tide at the diamond's standard port,
-- find the "tidal hour",
-- interpolate the rate,
-- find the resultant vector
- Add all the vectors to give a CTS.
- Do the whole thing again to get hourly EPs to make sure the whole leg is conducted in the water.
Ideally now you do do it again for each of the next twelve hours to find the shortest passage time. Try it for slightly different boat speeds...
Conceptually simple, but very time consuming and error-prone. A PC will march through the spherical-surface trig sums at millions a sec.
Here is were the program comes in.
Tidal data for standard ports is imported (by the user) from WXtide32 and users enter the tidal diamond data from their charts: the more diamonds you add the better the estimates. For cross channel one at each end and one in the middle seems to work. Coasting along the south coast one at each headland and one in between each seems to work well. A "library" of diamonds is built up over time.
"Routes" (of "legs") can be imported from Seaclear or typed in and the
resulting hourly EPs can be exported to Seaclear (as a "track").
Here is a route from Dover to Newhaven (imported from Seaclear and some of the default WP names have been left unedited)
and the whatif (next 12 hours)
Clearly the time to leave is around 16-17h
Here is the leg detail showing estimated duration of each leg and CTS. Also show is the optional expansion to show which hours of which diamonds the leg estimates are based on.
Finally you can import the hourly EPs back into Seaclear to display them. This is a different example, which shows the EP track more clearly than when coasting.
Its very early days yet. I needed work up a "real" program in C# in a Microsoft-free development environment, for reasons unrelated to sailing, and I knew I would do it better if it was interesting.
There are systems which do this but they are expensive and depend on getting new "packs" for new cruising areas. This scheme depends on data entered by the user, from charts.
Questions for the panel are:
Is this sensible?
Would folks pay modest sums of money for it?
