[LTER-mcm-pi] wait... there IS more variability in streamflow since 2002!

[John Barrett]

OK, so as a total block of time. What about the variance within a flow
year. Has that changed over time too? Basically the distinction is between
seasonal and inferannual variability.
On Aug 16, 2015 6:17 PM, "Michael Gooseff" <michael.gooseff at colorado.edu>
wrote:

> It’s the annual variance of the total flows.  So, we grouped the total
> annual flow volumes from 1990-2001 and from 2003-2013 then calculated
> variance on both groups.
>
> Mike
>
>
> From: John Barrett
> Date: Sunday, August 16, 2015 at 2:08 PM
> To: Michael Gooseff
> Cc: Berry Lyons, "mcm-pi at lternet.edu"
> Subject: Re: [LTER-mcm-pi] wait... there IS more variability in
> streamflow since 2002!
>
> Really cool mike. That's certainly the general impression.
> Is the increase in variance a within year phenomena or interannual?
> On Aug 16, 2015 4:56 PM, "Michael Gooseff" <michael.gooseff at colorado.edu>
> wrote:
>
>> Hi all,  (greetings from Toolik)
>>
>> I wanted to follow up on a conversation we had a few months ago about the
>> variability in the MDVs since 2002.  Recall that Chris shared with us a
>> spreadsheet of annual flow volumes from the Fryxell basin streams, and that
>> the analyses suggested that *variance had gone down in the 2003-2013
>> period (compared to 1990-2001)*.  My recollection is that this sort of
>> took the wind out of our sails on that discussion.
>>
>> I revisited these analyses in the past week in trying to get some writing
>> and a figure together for the Bioscience paper that Andrew is leading and
>> Chris and I revised the analyses -
>>
>>    1. We fixed an error we found in the calculation of % change of
>>    variance from pre- to post-flood.
>>    2. We took out the 2001-02 flood year (to look at ‘before’ and
>>    ‘after’, we argue that we shouldn’t lump the flood year in either if we
>>    postulate that it is a turning point)  —> *doing these 2 steps alone
>>    resulted in an increase of variance of 185% in the post-flood vs pre-flood
>>    period**…*
>>    3. *However*, we had several seasons in each period of poor or no
>>    flow estimates because of technical problems with gauges or controls (3 in
>>    the pre-flood period (~4%), 7 in the post-flood period (~7%)).  Chris came
>>    up with some nice correlations to fill in these gaps (no relationship has
>>    an R2 < 0.70), and *the result is an increase in variance of 48% in
>>    the post-flood period vs. the pre-flood period*.  This is (I think)
>>    our best analysis of the flow records and indicates increased variance in
>>    the recent decade.  (most updated figure inserted below with snip of text)
>>
>> So, *I’d like to suggest that we revisit the notion that our system is
>> becoming more variable as we emerge from the cooling trend as a potential
>> theme for the renewal*.  What are some of the other ecosystem parameters
>> we might evaluate for similar changes?
>> —> JP showed us UW PAR data last week that looks like it could be more
>> variable, but will be hard to evaluate with the gaps
>> —> Soil or Air temperature analyses?
>> —> Soil moisture analyses from a long-term plot?
>> —> chl-a in lake photic zone?
>>
>> If we can demonstrate change in ecosystem temporal trends (the synthesis
>> paper mega plot) and increased variance of some drivers, I think we have an
>> easy sell on the motivation for the proposal…
>>
>> Talk to you all on Thursday.
>>
>> Cheers
>> -Mike
>>
>>
>> *Snip of text:*  [in the pre-flood period] mean annual streamflow was
>> 0.85 x 106 m3 with a standard deviation of 0.93 x 106 m3 and a variance
>> over this period of 0.86 x 109 m3.  The highest annual flows on record
>> were observed in the 2001-02 flow season (total streamflow over 5 x 106 m
>> 3).  Since this ‘flood year’, mean annual discharge has increased
>> (2003-2013; 1.59 x 106 m3), as has the standard deviation (1.13 x 106 m3)
>> and the variance (1.28 x 109 m3, an increase of 49%).
>>
>>
>> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
>> Michael Gooseff, Associate Professor
>> Institute of Arctic and Alpine Research
>> Civil, Architectural, and Environmental Engineering
>> University of Colorado
>> Boulder, CO 80309-0450
>>
>> email: michael.gooseff at colorado.edu
>> web: http://goosefflab.weebly.com
>> phone: 303.735.5333
>>
>>
>> _______________________________________________
>> Long Term Ecological Research Network
>> mcm-pi mailing list
>> mcm-pi at lternet.edu
>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.lternet.edu/pipermail/mcm-pi/attachments/20150816/b12fd11d/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: A5029176-02CE-4C56-8D60-60E7C7B196AD.png
Type: image/png
Size: 823865 bytes
Desc: not available
URL: <http://lists.lternet.edu/pipermail/mcm-pi/attachments/20150816/b12fd11d/attachment-0001.png>