Aug 6 - Aug 14

[eeholmes]

Anthony

• update the coast angles to the latitudes off Morocco
• run Williams code on the Morocco data
• play with the daily data for the N America coast --> run through William's algorithm --> 45deg N to 26deg N
• Make a panel plot of the different latitudes
• Write a function to compute the n-day average.
• Write a function to compute the first day and last day above a threshold in a year

William

• Detrend algorithms. What to do if there are stepwise changes or hump shaped changes.
• Expand data along the Peruvian coast

Eli

• Look into daily or 6-hr FNMOC pressure data for long periods. NASA Earthdata.
• Aug 15-18: Find some folks to give a presentation to.

[eeholmes]

@XuweiyiChen
Paper on using log-periodic power-law to detect critical breaks in time series. Might be of interest?
https://www.scirp.org/journal/paperinformation.aspx?paperid=80887

[eeholmes]

@XuweiyiChen Baseline Wander is also a problem in EKG analysis. Lots of papers on methods for removing that.

• https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4678436/
• Or maybe a simple cubic spline is fine. Just don't want it to wiggle too much https://link.springer.com/chapter/10.1007/978-3-030-82014-5_34

[eeholmes]

Eli's update

This paper has a link to daily sea level pressure which they use to calculate the Ekman upwelling. Same ideas as the Ekman upwelling index that we have been using. Just a different source for air pressure (not FNMOC).

Paper: https://agupubs.onlinelibrary.wiley.com/doi/10.1002/2016JC012624
Data: https://disc.gsfc.nasa.gov/information/glossary?title=MERRA-2
How to download the data: https://disc.gsfc.nasa.gov/information/howto?keywords=MERRA-2&page=1
A Python notebook to show how: https://docserver.gesdisc.eosdis.nasa.gov/public/project/notebooks/How_to_calculate_and_plot_wind_speed_using_MERRA-2_wind_component_data.ipynb

[eeholmes]

William's update

Monthly data

Red points is where we are calculating

Detrend method

x(t) - x(t-1)

Just the yearly cycle

www.pfeg.noaa.gov/products/PFEL/modeled/indices/upwelling/SA/upwell_menu_SA.html

[eeholmes]

Anthony update

Anthony

Created a function

• sum over the period of upwelling / # days in that period
• start of period is when the 7-day average is first over a threshold
• end of period is when the 7-day average is below the threshold

Homogenization function

Paper said that more homogenization. up to 1950 to 2009 using CMiP5 historical simulations

• roughly a metric of the variance across the latitudes

Daily data for California homogenization trend 2017 to 2020

Morocco 1982 to 2022

In mid-2010s, less homogenized. In most recent years, less homogenized.

presentation

1. The nature paper suggests that homenization in intensity and duration will decrease
2. Your results show that it is more complex and that there is a northern and southern regions in the Canary area (off Morocco)