[prev in list] [next in list] [prev in thread] [next in thread] 

List:       postgis-users
Subject:    Re: [postgis-users] Spheroid Question
From:       Stephen Woodbridge <woodbri () swoodbridge ! com>
Date:       2005-06-28 13:51:15
Message-ID: 42C155D3.3080000 () swoodbridge ! com
[Download RAW message or body]

Markus Schaber wrote:
> Hi, Stephen,
> 
> Stephen Woodbridge wrote:
> 
> 
>>>>the_geom && expand(setSRID(makepoint(-72.607912,
>>>>42.058052),4326),0.179664)  AND
>>>
>>>[Meters per degree]
>>
>>You are correct that degrees per meter increases, but we are concerned
>>with meters per degree in this example and being the inverse they
>>decrease. So by using the value at the equator, we over sample as we
>>move toward the poles which is ok, unless it creates a performance
>>problem in which case you will have to expand the bbox x and y as you
>>suggest.
> 
> 
> AFAICS, you give an degree value 0.179664 to expand in your example, and
> this makes the bbox too narrow when going to the poles.

Hi Markus,
Oops, Right you are! This gets very confusing without a picture and 
numbers to validate the model. So somewhere very near the pole 1 degree 
of longitude == 1 meter and closer to the pole you need multiple degrees 
of longitude and at the pole it is a divide by zero and at the equator 
you need some fraction of a degree (0.179664) to equal a meter.

So like I said, user be aware of the limitations.

It might be nice to have a function or three that deals with these issues.

-Steve

[prev in list] [next in list] [prev in thread] [next in thread]