I have a list of GPS measurements for a point for a period of about 2 years. They are latitude, longitude and altitude. I want to generate the time series for the east, north and up components. Do I just need to convert the geographic coordinates to UTM coordinates or should I do something else?
I gather that you want to convert your spherical coordinates (ie/lat-long) to grid coordinates. There are various projection systems designed to do that. UTM is one. Whether UTM is the most appropriate for your needs depends on where you are and what you are doing with the results.?ÿ
Note that conversion of the lat/long to grid is a separate function from that of converting the heights from ellipsoidal to orthometric.?ÿ?ÿ
If it's only one point I would suppose you could draw the actual movement to true scale. Just make the y-axis represent seconds of arc/altitude, not needing conversion. The x-axis could be time, and depending on the frequency of the observations (daily? weekly? monthly?) could be spaced at 0.2 inch/5 mm for weekly data to fit on an 11"x17" ledger sheet.
A graph with red (northing), blue (easting), and green (altitude/elevation), for example, would do nicely.
The original poster wants to transform latitude, longitude and heights (LLh) to North, East and Up. What he wants are 3D values in order to generate a time series of differences.
I suggest reviewing http://www.mygeodesy.id.au/documents/LOCAL%20HORIZON%20COORDINATES.pdf
At this time, I cannot recollect any Internet-based tool that accepts as input LLh and returns ENU.
The paper linked above nicely describes the issue and shows how to do the rotations. Coding the problem is straightforward in Matlab or similar tools.
The paper is by Deakin. I highly recommend his site as a source for well-written papers on geodesy and related subjects.
FWIW, the NGS Invers3D tool does provide results in both DXDYDX and DEDNDU. I also add a graphic from the Deakin paper referenced in my previous post.
?ÿ
Since he wants a time series, there is no need to convert ellipsoid height to orthometric. Both change by the same amount
Isn't latitude, longitude and height already in a form of xyz coordinate?
Seems like needless calculation when the answer is right there.
You have to be careful here. What is the basis of Lat and Long?
?ÿ
Geodetic or Geocentric? You have to know what you're dealing with so you can make the proper conversion.
?ÿ
Latitude and Longitude are different lengths which vary by latitude (latitude much more slowly, but it does vary). If one wants to know how much something has moved, they are not as useful to plot directly.?ÿ
One can upload a file of lat/long to NCAT and get grid coordinates returned.
?ÿ
To answer the original question...just convert to grid (UTM, SPC, etc), nothing more is required.?ÿ
Geodetic or Geocentric?
That isn't going to matter for most projects as long as you stay consistent.
Depending on the Latitude, Longitude and Altitude? precisions he has, E, N and Up may just plot as an imprecise blob. We do need further explanation of what he calls Altitude.
Paul in PA
Pending clarification from the OP, I interpret his reference to ??altitude? to mean height. Also, his reference to UTM coordinates as a potential solution makes he wonder.
If the OP has the output from a GPS-based survey expressed in latitude, longitude and ellipsoid height and wants to transform the results to a more understandable NEU, he should use the Deakin paper.?ÿ
IF he has the vector information from a reference/base station to his site/s, he can transform the DXDYDZ information to DNDEDU. I would prefer this approach.?ÿ
Some have argued that the DXDYDZ values are sufficient. As the OP wants differences in terms of North, East and Up an additional step is required. Take a look at the output of the NGS tool FORWARD3D:
This output is based on a forward azimuth, zenith and ellipsoid distances observed at the ??from? station. You can calculate the NEU differences from the ??from? station using simple trigonometry. ?ÿ?ÿ
There are lots of folks way smarter than me, if they can look at the DXDYDZ values alone and break the differences into NEU, more power to them. For me, the NEU differences are much more relevant when trying to understand what is happening to a site.?ÿ
As I do not know what the OP intends to monitor nor do I know if he has other data sources (e.g. differential leveling in NAVD88 or even NGVD29), I encourage him to consult with his NGS Regional Advisor see:?ÿ https://www.ngs.noaa.gov/ADVISORS/?ÿ
Awaiting correction or enlightenment,
?ÿ
DMM
I have computed several ??time series? projects both for my private use as well as a couple of client projects. As with everything else, there is more than one way to skin this cat, and the choice of skinning technique (chili verses fur coat for SWMBO) depends A LOT on what you hope to determine.
The Lat/Lon/Height to N/E/U computations are the easy part!
The metadata supporting the input geographic coordinates (Latitude/Longitude/Height) of the subject station is the potential WILDCARD in the recipe (GIGO).
Without knowing the WHY, it's pretty hard to suggest a HOW.
@bill93 right
My line of thought was when receiving data from outside sources. You better know which type of Lat Long you're using.
Also, you need to be educated enough to understand what your collector is giving you vs. what you're plugging into it.
