Hi Jason,

I briefly looked at the Python code but was struggling to parse it (I mostly work in R so coding in python is a stretch for me) and unfortunately have never used nor have access to MATLAB. I will definitely reach out to the authors and see if they might consider some updates and/or provide some insight into running the code on updated data. Incorporating it into MGET in the longer term sounds great, so definitely keep me posted on that process and I will let you know if I have any luck with implementation in the shorter term.

Best,

Julia

On Tue, Sep 30, 2025 at 11:41 AM Jason Roberts <> wrote:

Hi Julia,

I have not looked at Xing's code, nor have replicated their algorithm. Sorry I can't be of more help there! Did they have Python code? When I looked originally I found some MATLAB code (here). It looked like it might be straightforward to run, insofar as you just provided an SST image and no other parameters. But there was only source available for the main algorithm, which called some helper functions that were only available as compiled MATLAB libraries (.mexa64 files).

Have you tried contacting the authors? Perhaps they could bring their archive up to date, or maybe give some suggestions about the code.

I did port MGET's MATLAB interoperability functionality from Python 2.x to 3.x. So, in principle, we could integrate the Xing algorithm into MGET, which might make it easier to run on other SST datasets. However, the compiled MATLAB libraries would likely be a problem. Unless we could get the source code for those, I don't know if it would work. Our own lab needs fronts data for 2024-2025 for one of our own projects, so I do have a justification for working on it under an existing grant. But this is not something I could consider until 2026.

Jason

On 9/26/25 14:35, Julia Gulka wrote:

Hi Jason,

Thanks so much for the quick response! That's interesting to hear that you've had better results with the Canny algorithm - everything I've read implies that C-C is less sensitive to parameter selection. Do you think it performs better specifically for the U.S. Atlantic or more generally? I did come across the Xing et al. algorithm and am intrigued. I am looking to associate our gannet tracking data from Project WOW to these environmental covariates and the timescale of the data is 2023-2025. The Xing et al dataset only goes up to 2023 which is why I am (albeit hesitantly) needing to derive my own front data. I was struggling to make sense of the python code from the authors so would be curious to know if you have replicated this front detection algorithm or just used the existing dataset? Any additional guidance would be greatly appreciated!

Best,

Julia

On Fri, Sep 26, 2025 at 2:10 PM Jason Roberts <> wrote:

Hi Julia,

Nice to hear from you. I hope to eventually port the C-C tool but am taking a break from MGET work for the next few months to finish some other projects. The C-C tool has not as been as high priority for us because we've had what I believe are better results from the Canny algorithm. More recently, we've also been using Xing et al.'s algorithm and dataset. I still think there is value in porting C-C to Python 3 / ArcGIS Pro, but the soonest I'd get to it would probably be January. Apologies I don't have a faster timeline!

Best,
Jason

On 9/26/25 11:19, Julia Gulka wrote:

Hi there,

I see that you have been working to get the MGET Toolbox into ArcGIS Pro and it looks like only some functions have been brought over from the original desktop version. In particular I am interested in the Cayula-Cornillion front detection capabilities and wondering if there are plans to incorporate this functionality and if so on what timeline?

Thanks,

Julia

--

Biodiversity Research Institute

Julia Gulka Associate Director Center for Research on Offshore Wind and the Environment  207-839-7600 x303  802-522-8750    https://briwildlife.org

---

BRI is a nonprofit ecological research group based in Portland, Maine, conducting innovative wildlife science worldwide.
This email may contain confidential or proprietary materials for the sole use of the intended recipient. If you are not the intended recipient, any review, use, distribution, or disclosure is strictly prohibited.

To unsubscribe from this list, visit: https://lists.nicholas.duke.edu/sympa/ca/initiate_unsubscribe/mget-help/jason.roberts%40duke.edu

--

Julia Gulka

Marine Biologist

Biodiversity Research Institute
276 Canco Rd., Portland, Maine 04103

(207) 839-7600 x303

www.briwildlife.org

--

Julia Gulka

Marine Biologist

Biodiversity Research Institute
276 Canco Rd., Portland, Maine 04103

(207) 839-7600 x303

www.briwildlife.org