[{ALLOW edit EISMainUsers}]
[{ALLOW view Anonymous}]
!!!Procedure for obtaining SDO cutouts from the JSOC service

SDO cutouts (both AIA and HMI) can be obtained very quickly through the JSOC. In this article I illustrate how cutouts corresponding to a specific EIS raster can be obtained. Note that the method is generally useful and does not apply specifically to EIS: all you need is a field-of-view size and a pointing location.

I consider an EIS active region observation that began at 02:30 UT on 15-Apr-2011. The raster took 62 mins to complete. We would like AIA 193 cutouts at 1 min cadence for the period 02:30 to 03:32, and HMI magnetograms at 45s cadence for the same period.

Note that Lockheed also provide a cutout service, but this is much slower than the JSOC and does not return HMI data.


!!Get EIS pointing and FOV size

The wiki page [Obtaining pointing information from EIS data|EISPointing] explains how to obtain the center and FOV of an EIS raster, so you can follow the instructions there. You should get values of (xcen,ycen)=(199,305), and (fovx,fovy)=(120,160).

!!Request AIA 193 cutouts

Go to the [JSOC exportdata webpage|http://jsoc.stanford.edu/ajax/exportdata.html]. In the 'RecordSet' box, type:

{{{
aia.lev1_euv_12s[2011.04.15_02:30/62m@1m][193]
}}}

Click on the 'Recount' button and make sure a non-zero number of Records are found.

For 'Method', choose 'url-tar' from the drop down list. You will see a new option called 'Processing' that appears. Tick the box.

Within the processing options, select 'im-patch - Extract sub-frame'. A new box appears with various input widgets, some of which will be automatically filled in. 

Make sure 'Tracking' is ticked, and also tick the 'Register' option.

For the other fields, type:

LocUnits: 'arcsec from center'\\
T_Ref: '2011-04-15T03:01:02Z' (choose time from middle of raster)\\
X: 199\\
Y: 305\\
BoxUnits: pixels\\
Width: 300 (some contingency has been factored in here)\\
Height: 400 (some contingency has been factored in here)

For 'Protocol' leave it as 'FITS'. Type in your name and e-mail, and then click on 'Submit Export Request'. ''If you're doing this for the first time, then you will have to register.'' After a couple of minutes wait, click on 'Submit status request' and you should see a link to a tar file containing your data (it may take longer for large data-sets).

Note that cutouts will be rice-compressed images (there used to be an option to download uncompressed images, but this has been removed). See below for how to read compressed images quickly.

Users should be aware that the AIA de-spiking routine (which is automatically applied to all files) can remove real data, which particularly affects small-scale, dynamic phenomena. The spikes can be recovered, however, so please check the page [re-spiking AIA images|AIARespike].

!!Request HMI cutouts

Repeat the above procedure, except that in the RecordSet box you should do:

{{{
hmi.M_45s[2011.04.15_02:30/62m@45s]
}}}

For the field of view, you need to account for the difference in plate scale between AIA and HMI, so multiply by 1.2:

Width: 360\\
Height: 480

!!Downloading cutout images from IDL

After you have submitted your cutout request to the JSOC you will receive a request_id of the form 'JSOC_20161007_486'. You can check the status of your request from IDL by doing:

{{{
IDL> sdo_jsoc_check_status, request_id
}}}

which will check to see if the tar file is available for downloading. If it is, then you should see a message like:

{{{
The data are available at the webpage:
   http://jsoc.stanford.edu/SUM9/D862749672/S00000
}}}

You can download and unpack the tar file to your current directory by doing:

{{{
IDL> sdo_jsoc_check_status, request_id, /download
}}}

Note that the tar file is automatically removed after unpacking.


!!Reading cutout images

A cutout image can be read with:

{{{
IDL> read_sdo, file, index, data, /use_shared_lib
}}}

The /use_shared_lib keyword is highly recommended as it greatly increases the speed at which rice-compressed images are read. It may not work on all computers though (Windows?).

!!Further information

Shea Hess Webber and Peter Young have written a [document|https://pyoung.org/gsfc/JSOC%20Export%20Process.pdf] giving more information about using the JSOC service.

!!Comments

''Why use pixels instead of arcsec for the FOV size?''\\
I found erratic results when using arcsec - the HMI FOV would end up with a different size to AIA.

''Why use the register option?''\\
This performs sub-pixel interpolations so that when you make a movie from the cutouts you do not see the occasional pixel jumps that otherwise would be seen. This isn't recommended if you plan to re-spike your cutouts.

''What if I want multiple AIA filters?''\\
Instead of giving {{{[193]}}} in the RecordSet box, give e.g. {{{[171,304,335]}}}.

''How do I get AIA UV images?''\\
For these you need to give 'aia.lev1_uv_24s' in the RecordSet box.

''How do I get HMI 45s intensitygrams, or HMI 720s magnetograms?''\\
In the RecordSet box, use 'hmi.Ic_45s' and 'hmi.M_720s'.

''I want to respike my cutouts. Can I get spike cutouts?''\\
Please check the page [re-spiking AIA images|AIARespike].

''What about HARPs? How do I extract information for these?''\\
A HARP is a HMI Active Region Patch, i.e., a cutout of HMI data that has been processed to yield, e.g., the vector magnetic field. See the separate page on [Extracting HARP data from the JSOC|JsocHarp].

''Can I get header information from JSOC without downloading data?''\\
Putting the following URL into your browser will give a text page with the specified keywords (T_OBS, WAVELNTH, EXPTIME and NSPIKES) given as columns.

{{{
http://jsoc.stanford.edu/cgi-bin/ajax/show_info?ds=aia.lev1_euv_12s[2011.04.15_02:30/62m@1m][193]&key=T_OBS,WAVELNTH,EXPTIME,NSPIKES
}}}