And here again is the curve we discussed on Tuesday from Zhao and Running for the same thing from 2000-2009 (also green curve):
(Note that the earlier paper also involved the same University of Montana group, along with other collaborators).
On Tuesday, I offered my opinion that the second set of data is not a statistically adequate basis to say that there has been a "reduction" in NPP over that timeframe.
A more involved question is this: can we look at these two sets of data together and confidently infer at least this much: the NPP trend is getting worse than it used to be?
In the interests of time, I will offer an analysis tomorrow. In the meantime, readers are encouraged to offer their opinions...
11 comments:
It's worth noting that a number of other groups have been unable to replicate the Nemani et al. results (see, for example, Bunn and Goetz 2006, which whom I have no affiliation).
Given that, the question that occurs to me, looking at the much-longer Nemani record, is something like: "given the observed variability in the 1982-1999 data, how many years of data would we need to confidently observe a 0.5Gt decline, as put forth by Zhao and Running?" In other words, a power calculation.
Ben:
Are you referring to this paper? It is a study only of the areas north of latitude 50 degrees...
Yes. Sorry, I didn't mean to imply that Bunn and Goetz was a global study--just that (from what I understand from my remote sensing friends) the results of the Nemani paper you cited have proved difficult to replicate.
At any rate your idea of using the older Nemani data with the new Zhao results seems interesting, if only to put them in a longer context.
Ben:
Thx - interesting! I had wondered about how much to trust the underlying analysis that gave rise to the time series, given gross error in drawing conclusions from the time series... But there's not much to be done about that - certainly not by me as a non-specialist.
A full analysis of the two series with treatment of the autocorrelation is pretty complex I realize. The Nemani et al series is clearly not white. In addition to worrying about the potential for multi-decadal oscillations in the climate input to vegetation there's clearly potential for unit-root random-walk type stuff - if you have a good/bad year for NPP that leaves more/less plant tissue in place at the beginning of the next year, which can clearly influence next year's NPP. And only 18 years of data...
I think the basic logic is following:
climate change already is and further WILL exacerbate weather extremes - so this, in global average, should further counteract CO2 fertilization/longer season largely positive effects.
further increase in global temperature wil drive respiration up (since this is exponentialy dependent upon temp.) while supress photosynthesis...
I think there is no reason whatsoever to expect the situation might stay as it is today and not getting much much worse (see the increasing intensity/and or frequency of heatwaves, droughts, floods, pest and fungi outbrakes, wildfires, increasing human population, deforestation etc)
So my bet is that in 10 years we have statistically significant decrease in NPP anomaly...
Alex
NPP could or could not be a big deal globally. However lots of flora and fauna is already stressed and unable to adapt at current latitudes due to increased temperatures and dryness. So all things being equal you measure NPP but all things not being equal, plants go extinct as they can't adapt and then there are less plants and new ones don't grow as well up nortt, instead you get wild fires in Tundra, for example.
May be of interest: CO2 measurements. Graph at 5.1 has "natural sinks" taking up the slack when emissions don't register as observed CO2; how this differs from measuring NPP I don't know. Latter is more detailed? But why not flesh out info with numbers from former, if only for contrast?
Terrestrial NPP accounts for about half of the global production, with the other half coming from oceanic phytoplankton. In a widely reported paper in Nature last month, Boyce et al demonstrated that global phytoplankton production in the open ocean declined over the last century, with the mean rate of decline being about 1% per year.
Regional trends were correlated with climate indices. Sea surface temperature was the single strongest physical correlate, with rising temperature being related to declining production over most of the global ocean, probably through the effect of surface temperature on mixing and nutrient supply. The polar oceans showed an opposite trend.
The authors assembled a truly impressive data set to arrive at their conclusions. Blended from a number of different observational types, it contained almost 450,000 chlorophyll measurements. Their statistical analysis makes the general decline in oceanic productivity clear, but I'm disquieted by the shape of many of the regional trend curves and their observation that, in a data set going back to 1899, the chlorophyll maxima in all ten ocean regions has occurred since 1996, the majority in the last four years.
Hi,
also a graph from the paper Heimann and Reichstein (2008) published in Nature might be of relevance here: The 1. graph in their paper shows growh rates of CO2 together with ENSO oscillation (since 1958). When ENSO index is high (El Nino), global temperatures are higher and CO2 growth rates are higher (lower NPP) and vice versa. up to now the CO2 fertilization and longer growing season counteracted global climate disruption.
BTW I do not see how higher temperatures might be benefical for NPP when CO2 growth rates are higher were warmer years...
So to conclude, I fail to see how NPP might NOT decrease in the future, if the warming continues (and it will)...
link to that paper is here:
https://www.bgc-jena.mpg.de/bgc-mdi/uploads/Publ/Heimann_and_Reichstein_2008.pdf
And another one:
"1) The response of the ecosystem carbon balance to climate variability is highly skewed, i.e. large carbon losses are much more likely than large carbon gains during periods with climatic conditions deviating from normal,
2) climate variability and extremes contribute to a “slow-in fast-out” effect, where
several years of slow carbon uptake induced by gradual climate change can be undone by single extreme events,
3) the vulnerability of carbon pools on a per-area basis generally correlates with general pool size, but there are important ecosystem dependent exceptions,
4) there are strong interactions between different climate drivers,
5) legacy effects have to be considered for full accounting of the effects of climate variability on long-term carbon balance."
link here: https://www.bgc-jena.mpg.de/bgc-mdi/uploads/Publ/Reichstein_et_al_2009.pdf
Post a Comment