topic Re: Predicting a lake shape 6 years later in Spatial Statistics Questions

Predicting a lake shape 6 years later

MOHAMMEDHABBOUB — Tue, 22 Jan 2013 08:52:13 GMT

Hi there,

I'm looking forward predicting a lake shape 6 years later, in the same condition trends, I have 20 shapefiles describing the shape of the lake in a series of years.

and how could I make a the DEM to contribute because it will affect the shape of the lake

Could anybody give me a hint?

Re: Predicting a lake shape 6 years later

WilliamFarebrother — Mon, 28 Jan 2013 23:48:52 GMT

Hi Mohammed,

If you download a SRTM DEM or similar DEM for your lake region you could convert your polygon to a series of points (i.e. feature vertices to points) and then using the extract values to points (using the DEM as the raster source and the vertices points as the point feature) it will give you the elevation of the lake shoreline at each point. This would allow a crude estimation of any trends between each year. To predict future lake shorelines you could use predicted lake shoreline elevations and fill the DEM to that elevation effectively creating the predicted lake shoreline

Dakota_20

Re: Predicting a lake shape 6 years later

MOHAMMEDHABBOUB — Thu, 31 Jan 2013 16:34:17 GMT

Hi Mohammed,

This would allow a crude estimation of any trends between each year. To predict future lake shorelines you could use predicted lake shoreline elevations and fill the DEM to that elevation effectively creating the predicted lake shoreline

Dakota_20

That's nice but the problem is that the lake is shrinking not expanding and the DEM doesn't tell me the elevation under water so I can't get the elevation under water body !!

Re: Predicting a lake shape 6 years later

MarcoBoeringa — Thu, 31 Jan 2013 19:43:29 GMT

That's nice but the problem is that the lake is shrinking not expanding and the DEM doesn't tell me the elevation under water so I can't get the elevation under water body !!

The only reliable way in case of a shrinking lake would be to determine the bottom surface by means of sounding, for example using a multi-beam sonar, and than use that to fill up the gap in the DEM.

A crude alternative is to determine some weighted centre point of the smallest circumference, and than model the shrinkage in let's say each degree direction of 360 degree round based on your existing shapes, and than use that somehow as input for an ArcGIS spatial adjustment rubbersheeting operation to automatically define "displacement links".

But this isn't going to be an "easy" solution and modelling operation...