search

helmingstay / unm-r-programming

This is the permanent website for the University of New Mexico (UNM) R Programming Group, which met from 2012-2014.
0 stars 2 forks source link

Identify weather events #5

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
We would like some way to identify notable weather days.  MinPressure, MaxWind, 
and MaxGusts don't seem to tell much, or to be really noisy.

Attached is a script file (and new data file) to pull out the top X records of 
Mean.Wind (though you can use a different variable) for each of the weather 
stations and merge them into a single XTS, and then plot them.  Notice that 
high wind days are often seen at multiple stations.

I have 2 questions:

1.  What is the putative mechanism by which weather affects birds?  Is it 
primarily high winds, or rain/lightning, is temperature a part??

2.  What is the measure of "novelty" -- bird rareness/diversity that "responds" 
to storms/weather events?

Original issue reported on code.google.com by icos.atr...@gmail.com on 27 Mar 2012 at 11:47

Attachments:

GoogleCodeExporter commented 9 years ago 
1. The main reason we should see "novel" events after storms is because of high 
winds blowing the birds around/off course. Rain/lighting might be associated 
with fewer sightings, mostly because the birds/birders are staying in. Not sure 
what affect temp might have but I doubt there's that much variation in PR 
right? 

2. For novelty... what about comparing abundances of birds after events to 
their "baseline" or average abundance over the whatever period?

Original comment by sabrina....@gmail.com on 27 Mar 2012 at 3:10

GoogleCodeExporter commented 9 years ago 
I agree with Sabrina, and like her idea of comparing abundances of birds
immediately after events to their baseline/average over that month across
all years we have data for.

I'm pretty sure it really is the high winds that cause odd birds to show up
- birds caught flying by storms can get blown way off course by the winds.
There might also be a confusion factor for migrating birds, but I don't
know how much evidence there is for that.

Original comment by coereba....@gmail.com on 27 Mar 2012 at 3:16

GoogleCodeExporter commented 9 years ago 
I haven't looked at this code yet, but I made a wiki page for tropical cyclone 
events affecting PR. The pdfs I linked to have data from some weather stations 
for wind, rainfall, pressure. We should just pick some of these events to check.

http://code.google.com/p/unm-r-programming/wiki/TropicalCycloneEventsForPR2002_2
011

Original comment by mjo...@gmail.com on 27 Mar 2012 at 6:48

GoogleCodeExporter commented 9 years ago 
Problems... I'm trying to look at the code but having trouble with the xts 
package. I installed zoo and xts but when I try to library(xts) I get this 
error message:

Error in dyn.load(file, DLLpath = DLLpath, ...) : 
  function 'zoo_lag' not provided by package 'zoo'
In addition: Warning message:
package 'xts' was built under R version 2.13.1 
Error: package/namespace load failed for 'xts'

Is there something else I'm missing? Sorry I'm not being much use but I'm kind 
of stuck.

Original comment by sabrina....@gmail.com on 28 Mar 2012 at 4:03

GoogleCodeExporter commented 9 years ago 
Where there any errors when you installed zoo?  Does require(zoo) work?

Original comment by icos.atr...@gmail.com on 28 Mar 2012 at 4:14

GoogleCodeExporter commented 9 years ago 
I installed xts which automatically installed zoo. This code also requires the 
lattice and foreach packages.

I was able to make the plot.

Original comment by mjo...@gmail.com on 28 Mar 2012 at 4:28

GoogleCodeExporter commented 9 years ago 
Maybe uninstall xts and zoo, and then try to install just xts and see what 
happens

Original comment by mjo...@gmail.com on 28 Mar 2012 at 4:28

GoogleCodeExporter commented 9 years ago 
One thing to check -- do you have a recent version of R installed?

Original comment by icos.atr...@gmail.com on 28 Mar 2012 at 4:46

GoogleCodeExporter commented 9 years ago 
Does someone want to start making a poster in powerpoint?  This might be my 
least favorite part of poster-making -- laying out title, section headings, etc.

I've worked up a few figures (see attached) -- still working on them.  I 
haven't found much in wind vs. rare sightings.  We'll have some nice maps of 
bird diversity, though. I'll throw code up tomorrow.  I have committed 
everything to the repository -- see the "source" tab at the top of the page.

Original comment by icos.atr...@gmail.com on 28 Mar 2012 at 10:19

GoogleCodeExporter commented 9 years ago 
attached...

Original comment by icos.atr...@gmail.com on 28 Mar 2012 at 10:19

Attachments:

GoogleCodeExporter commented 9 years ago 
Last comment -- if you're curious, look at the .R files here:
http://code.google.com/p/unm-r-programming/source/browse/#git%2Fresearchday2012

Original comment by icos.atr...@gmail.com on 28 Mar 2012 at 10:21

GoogleCodeExporter commented 9 years ago 
Another question:  Ideas how we're going to print this? The e-mail guidelines 
that we got say that we're supposed to print by today. We could always go to 
kinkos but that'll be expensive. Does anyone known Rihannon West/ if there are 
other ways to print? I'm going to e-mail her and ask if there are any times 
available tomorrow

Original comment by sabrina....@gmail.com on 28 Mar 2012 at 2:23

GoogleCodeExporter commented 9 years ago 
Christian and I have keys to the printing room, so Christian will print it
whenever it's finished.  No need for an intermediary.

Natalie

Original comment by coereba....@gmail.com on 28 Mar 2012 at 2:30

GoogleCodeExporter commented 9 years ago 
good work

Original comment by sabrina....@gmail.com on 28 Mar 2012 at 2:32