From aleece.nanfito at bhp.com Thu Sep 5 07:38:28 2019 From: aleece.nanfito at bhp.com (Nanfito, Aleece) Date: Wed, 4 Sep 2019 21:38:28 +0000 Subject: [GPlates-discuss] Convert reconstructed points back to raster Message-ID: Hello - I am looking for some help with converting reconstructed points back to a raster in pyGPlates. I received some excellent guidance from this group a few months ago on how to convert a raster (paleo-position) to points and then un-reconstruct the points to present day, but I am missing the part on how to convert the output points back to a raster. I would like the new raster to be the same resolution as the input raster and to minimize any interpolation or extrapolation of the point data. Apologies if this was answered in responses to my first question, but I am struggling to identify this last step. I appreciate any input and thanks so much in advance! Best regards, Aleece [cid:image002.png at 01D5633F.2CF65BD0] Aleece Nanfito Geologist aleece.nanfito at bhp.com T +1 713 499 5488 M +1 402 995 1976 1500 Post Oak Blvd. Houston TX 77056 USA bhpbilliton.com ________________________________ This message and any attached files may contain information that is confidential and/or subject of legal privilege intended only for use by the intended recipient. If you are not the intended recipient or the person responsible for delivering the message to the intended recipient, be advised that you have received this message in error and that any dissemination, copying or use of this message or attachment is strictly forbidden, as is the disclosure of the information therein. If you have received this message in error please notify the sender immediately and delete the message. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image002.png Type: image/png Size: 860 bytes Desc: image002.png URL: From john.cannon at sydney.edu.au Thu Sep 5 15:58:45 2019 From: john.cannon at sydney.edu.au (John Cannon) Date: Thu, 5 Sep 2019 05:58:45 +0000 Subject: [GPlates-discuss] Convert reconstructed points back to raster In-Reply-To: References: Message-ID: Hi Aleece, If you're using GMT then perhaps 'nearneighbor' could be used: https://protect-au.mimecast.com/s/FlEtCANZvPiQ21l6cG9SXa?domain=gmt.soest.hawaii.edu You first need to store your points (and raster z values) in an xyz text file with 3 columns (longitude, latitude, z). Then pass that to 'neighneighbor': gmt nearneighbor -N4 -S<1.5 * grid_spacing> -I -R-180/180/-90/90 -G ...with '<...>' replaced with your own values. I find the search radius (-S) of 1.5 times grid spacing to work fairly well in general. The data points (at the 'from' time) are no longer uniform (like at the 'to' time), so some interpolation smoothing is inevitable. Regards, John From: GPlates-discuss On Behalf Of Nanfito, Aleece Sent: Thursday, 5 September 2019 7:38 AM To: gplates-discuss at mailman.sydney.edu.au Subject: [GPlates-discuss] Convert reconstructed points back to raster Hello - I am looking for some help with converting reconstructed points back to a raster in pyGPlates. I received some excellent guidance from this group a few months ago on how to convert a raster (paleo-position) to points and then un-reconstruct the points to present day, but I am missing the part on how to convert the output points back to a raster. I would like the new raster to be the same resolution as the input raster and to minimize any interpolation or extrapolation of the point data. Apologies if this was answered in responses to my first question, but I am struggling to identify this last step. I appreciate any input and thanks so much in advance! Best regards, Aleece [cid:image002.png at 01D56402.CB9772B0] Aleece Nanfito Geologist aleece.nanfito at bhp.com T +1 713 499 5488 M +1 402 995 1976 1500 Post Oak Blvd. Houston TX 77056 USA bhpbilliton.com ________________________________ This message and any attached files may contain information that is confidential and/or subject of legal privilege intended only for use by the intended recipient. If you are not the intended recipient or the person responsible for delivering the message to the intended recipient, be advised that you have received this message in error and that any dissemination, copying or use of this message or attachment is strictly forbidden, as is the disclosure of the information therein. If you have received this message in error please notify the sender immediately and delete the message. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image002.png Type: image/png Size: 2555 bytes Desc: image002.png URL: From Christian.Heine at shell.com Thu Sep 5 16:58:40 2019 From: Christian.Heine at shell.com (Christian.Heine at shell.com) Date: Thu, 5 Sep 2019 06:58:40 +0000 Subject: [GPlates-discuss] Convert reconstructed points back to raster In-Reply-To: References: Message-ID: Hi Aleece, seconding John's suggestion and also in my view the easy solution. Use GMT (https://protect-au.mimecast.com/s/-9WBCANZvPiQ2J5WiGnJD5?domain=generic-mapping-tools.org) to pre/postprocess the points. The way I would do this is to call the GMT routines in a pygplates script either through os.subprocess() or even try to test the version of pygmt (GMT python api, https://protect-au.mimecast.com/s/LP79CBNZwLiGBJN4S6zHgx?domain=pygmt.org). Slightly different high level workflow compared to the one John proposed: 1. inputgrid -> grd2xyz -> xyz grid 2. use python or bash magic to associate points with plate IDs through "assing plate id function" 3. reconstruct & export point positions * potentially generate masks (using plate polygons in case you don't have full global/regional coverage due to point disappearing) 4. use GMT's blockmean/median/mode & surface with appropriate parameters re grid spacing and block sizes and gridding algorithms Cheers, Christian From: GPlates-discuss On Behalf Of Nanfito, Aleece Sent: Wednesday, 4 September 2019 11:38 PM To: gplates-discuss at mailman.sydney.edu.au Subject: [GPlates-discuss] Convert reconstructed points back to raster Hello - I am looking for some help with converting reconstructed points back to a raster in pyGPlates. I received some excellent guidance from this group a few months ago on how to convert a raster (paleo-position) to points and then un-reconstruct the points to present day, but I am missing the part on how to convert the output points back to a raster. I would like the new raster to be the same resolution as the input raster and to minimize any interpolation or extrapolation of the point data. Apologies if this was answered in responses to my first question, but I am struggling to identify this last step. I appreciate any input and thanks so much in advance! Best regards, Aleece [cid:image002.png at 01D563BF.24A6C430] Aleece Nanfito Geologist aleece.nanfito at bhp.com T +1 713 499 5488 M +1 402 995 1976 1500 Post Oak Blvd. Houston TX 77056 USA bhpbilliton.com ________________________________ This message and any attached files may contain information that is confidential and/or subject of legal privilege intended only for use by the intended recipient. If you are not the intended recipient or the person responsible for delivering the message to the intended recipient, be advised that you have received this message in error and that any dissemination, copying or use of this message or attachment is strictly forbidden, as is the disclosure of the information therein. If you have received this message in error please notify the sender immediately and delete the message. -- Christian Heine, Ph.D. Senior Geodynamicist | Structural Geology & Carbonate Team Specialist Geosciences PTD/E/F Shell Global Solutions International B.V. Lange Kleiweg 40, 2288 GS Rijswijk SIP: +31 70 447 5541 W: https://protect-au.mimecast.com/s/SXWDCE8kz9t7MojDfPdiTJ?domain=shell.com G: https://protect-au.mimecast.com/s/z14pCGvmB5i7GZ9MfkoPYI?domain=goo.gl -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image002.png Type: image/png Size: 9206 bytes Desc: image002.png URL: From t.j.m.vanderlinden at uu.nl Fri Sep 13 19:31:16 2019 From: t.j.m.vanderlinden at uu.nl (Linden, T.J.M. van der (Thomas)) Date: Fri, 13 Sep 2019 09:31:16 +0000 Subject: [GPlates-discuss] Generate Crustal Thickness Points in pyGPlates Message-ID: Hi everyone, Is the "Generate Crustal Thickness Points" from GPlates also available in pyGPlates or will it be in the near future? I would really like to use it in scripts. Otherwise is there a publication on the methods behind this feature? I looked into coding something similar myself, but ran into problems making a proper grid and with assigning plate-IDs. Hope you can help me! Best from Berlin, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From john.cannon at sydney.edu.au Fri Sep 13 21:03:24 2019 From: john.cannon at sydney.edu.au (John Cannon) Date: Fri, 13 Sep 2019 11:03:24 +0000 Subject: [GPlates-discuss] Generate Crustal Thickness Points in pyGPlates In-Reply-To: References: Message-ID: Hi Thomas, It should be available in the next pyGPlates release (sometime in the first half of next year). That release is aimed to bring deformation on par with current GPlates, so deformation-related things like generating crustal thickness points should get included. The crustal point locations are based on subdivision of a Rhombic Triacontahedron, with optional random perturbation of the positions. It's not perfectly uniform on the sphere but quite close. The source code file to look at is "src/maths/GeneratePoints.h": https://protect-au.mimecast.com/s/rhN2CJyp0qhJJLwoIVfcfG?domain=sourceforge.net . There are other algorithms like HEALPix (Simon Williams uses this) which has a library (but not on Windows). There's two separate aspects here, (1) the deforming topologies and their quantities like strain rate, and (2) reconstructing points in those regions. It's the latter aspect that you are probably interested in, and this is covered briefly in "GPlates: Building a Virtual Earth Through Deep Time" in the section "Reconstructing feature using topologies": https://protect-au.mimecast.com/s/g6b2CK1qJZtooxWkcv-Elr?domain=doi.org https://protect-au.mimecast.com/s/Eo8uCL7rK8tWWM92IPkOJu?domain=researchgate.net But it might be easier to wait until it's exposed via pyGPlates since moving these points through a deforming region is done by spatially interpolating triangulations of the deforming region as it evolves through time, and there's a few implementation gotchas to look out for. Regards, John From: GPlates-discuss On Behalf Of Linden, T.J.M. van der (Thomas) Sent: Friday, 13 September 2019 7:31 PM To: gplates-discuss at mailman.sydney.edu.au Subject: [GPlates-discuss] Generate Crustal Thickness Points in pyGPlates Hi everyone, Is the "Generate Crustal Thickness Points" from GPlates also available in pyGPlates or will it be in the near future? I would really like to use it in scripts. Otherwise is there a publication on the methods behind this feature? I looked into coding something similar myself, but ran into problems making a proper grid and with assigning plate-IDs. Hope you can help me! Best from Berlin, Thomas -------------- next part -------------- An HTML attachment was scrubbed... URL: From john.cannon at sydney.edu.au Fri Sep 13 21:19:27 2019 From: john.cannon at sydney.edu.au (John Cannon) Date: Fri, 13 Sep 2019 11:19:27 +0000 Subject: [GPlates-discuss] Convert reconstructed points back to raster In-Reply-To: References: Message-ID: Hi Aleece, Actually, if the points are uniform then it's better to use GMT 'xyz2grd' (instead of 'nearneighbor'). I just realized that I said your 'to' points are uniform so I should have recommended using 'xyz2grd'. Here I'm assuming you're generating your raster at the 'to' time using Simon's method covered here: https://mailman.sydney.edu.au/pipermail/gplates-discuss/2019-July/000748.html Regards, John From: GPlates-discuss On Behalf Of John Cannon Sent: Thursday, 5 September 2019 3:59 PM To: GPlates general discussion mailing list Subject: Re: [GPlates-discuss] Convert reconstructed points back to raster Hi Aleece, If you're using GMT then perhaps 'nearneighbor' could be used: https://protect-au.mimecast.com/s/QWUeCr8DLRtqqNy8i76WTx?domain=gmt.soest.hawaii.edu You first need to store your points (and raster z values) in an xyz text file with 3 columns (longitude, latitude, z). Then pass that to 'neighneighbor': gmt nearneighbor -N4 -S<1.5 * grid_spacing> -I -R-180/180/-90/90 -G ...with '<...>' replaced with your own values. I find the search radius (-S) of 1.5 times grid spacing to work fairly well in general. The data points (at the 'from' time) are no longer uniform (like at the 'to' time), so some interpolation smoothing is inevitable. Regards, John From: GPlates-discuss > On Behalf Of Nanfito, Aleece Sent: Thursday, 5 September 2019 7:38 AM To: gplates-discuss at mailman.sydney.edu.au Subject: [GPlates-discuss] Convert reconstructed points back to raster Hello - I am looking for some help with converting reconstructed points back to a raster in pyGPlates. I received some excellent guidance from this group a few months ago on how to convert a raster (paleo-position) to points and then un-reconstruct the points to present day, but I am missing the part on how to convert the output points back to a raster. I would like the new raster to be the same resolution as the input raster and to minimize any interpolation or extrapolation of the point data. Apologies if this was answered in responses to my first question, but I am struggling to identify this last step. I appreciate any input and thanks so much in advance! Best regards, Aleece [cid:image001.png at 01D56A78.0D331910] Aleece Nanfito Geologist aleece.nanfito at bhp.com T +1 713 499 5488 M +1 402 995 1976 1500 Post Oak Blvd. Houston TX 77056 USA bhpbilliton.com ________________________________ This message and any attached files may contain information that is confidential and/or subject of legal privilege intended only for use by the intended recipient. If you are not the intended recipient or the person responsible for delivering the message to the intended recipient, be advised that you have received this message in error and that any dissemination, copying or use of this message or attachment is strictly forbidden, as is the disclosure of the information therein. If you have received this message in error please notify the sender immediately and delete the message. -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: image001.png Type: image/png Size: 2555 bytes Desc: image001.png URL: