# AGN DP2 light curves

**URL:** <https://www.rubin.community/t/agn-dp2-light-curves/12717>\
**Category:** Science\
**Tags:** agn-variability, agnsc\
**Created:** [September 18, 2026, 8:56pm UTC](https://www.rubin.community/t/agn-dp2-light-curves/12717 "2026-09-18T20:56:25Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![mmarculewicz](https://avatars.discourse-cdn.com/v4/letter/m/a183cd/32.png) [@mmarculewicz](https://www.rubin.community/u/mmarculewicz)\
**Post date:** [September 18, 2026, 8:56pm UTC](https://www.rubin.community/t/agn-dp2-light-curves/12717/1 "2026-09-18T20:56:25Z")

</div>

Hi,

I was trying to use notebook examples to retrieve the data.  
I would like to follow up on the conversation we started: [DP2 photometry for AGN](http://www.rubin.community/t/dp2-photometry-for-agn/12672), which has some

The question is how do you usually treat the data and trust the photometry?

Rather than reliably recording the intrinsic variations of AGN, these light curves are showing some systematic errors in the photometric extraction, I am not sure how you usually treat and deal with them.

Are you also extracting light curves from some nearby comparison stars to check how the comparison stars behave and which one was used as a comparison star for a given AGN?

I would appreciate comments on extraction, reliable photometry, uncertainties, etc.

---

<div class="post-metadata">

**Author:** ![aetviita](https://sea2.discourse-cdn.com/flex002/user_avatar/www.rubin.community/aetviita/32/2659_2.png) [@aetviita](https://www.rubin.community/u/aetviita)\
**Post date:** [September 24, 2026, 2:41pm UTC](https://www.rubin.community/t/agn-dp2-light-curves/12717/2 "2026-09-24T14:41:17Z")

</div>

Hi,

We worked a bit on this topic in the context of our AGILE simulation paper: [https://ui.adsabs.harvard.edu/abs/2026A%26A...710A..56V/abstract](https://ui.adsabs.harvard.edu/abs/2026A%26A...710A..56V/abstract)

Sec. 10.2 describes some of the quality cuts that we found to be significant.

Note that some of these might already be redundant with the updated versions of the LSST Science Pipelines, into which I did not look in detail yet.

The general process of cleaning the data was a simple iterative one where we tried to look for properties that correlate with catastrophic outliers. Some usual suspects are low airmass (especially for bright sources), bad photometric/astrometric solution, vicinity of bright stars, and measurements near detector CCD gaps.

Hope this helps!