As I work on finishing our forcing/feedback paper for re-submission to Journal of Geophysical Research – a process that has been going on for months now – I keep finding new pieces of evidence in the data that keep changing the paper’s focus in small ways.

For instance, yesterday I realized that NASA Langley has recently updated their CERES global radiative budget measurement dataset through 2008 (it had previously ran from March 2000 through August 2007).

I’ve been anxiously awaiting this update because of the major global cooling event we saw during late 2007 and early 2008. A plot of daily running 91-day global averages in UAH lower tropospheric (LT) temperature anomalies is shown below, which reveals the dramatic 2007-08 cool event.

I was especially interested to see if this was caused by a natural increase in low clouds reducing the amount of sunlight absorbed by the climate system. As readers of my blog know, I believe that most climate change – including “global warming” – in the last 100 years or more has been caused by natural changes in low cloud cover, which in turn have been caused by natural, chaotic fluctuations in global circulation patterns in the atmosphere-ocean system. The leading candidate for this, in my opinion, is the Pacific Decadal Oscillation…possibly augmented by more frequent El Nino activity in the last 30 years.

Now that we have 9 years of CERES data from the Terra satellite, we can more closely examine a possible low cloud connection to climate change. The next figure shows the changes in the Earth’s net radiative balance as measured by the Terra CERES system. By “net” I mean the sum of reflected shortwave energy (sunlight), or “SW”, and emitted longwave energy (infrared) or “LW”.

The changes in the radiative balance of the Earth seen above can be thought of conceptually in terms of forcing and feedback, which are combined together in some unknown proportion that varies over time. Making the interpretation even more uncertain is that some proportion of the feedback is due not only to radiative forcing, but also to non-radiative forcing of temperature change.

So the variations we see in the above chart is the combined result of three processes: (1) radiative forcing (both internal and external), which can be expected to cause a temperature change; (2) radiative feedback upon any radiatively forced temperature changes; and (3) radiative feedback upon any NON-radiatively forced temperature changes (e.g., from tropical intraseasonal oscillations in rainfall). It turns out that feedback can only be uniquely measured in response to NON-radiatively forced temperature changes, but that’s a different discussion.

The SW component of the total flux measured by CERES looks like this…note the large spike upward in reflected sunlight coinciding with the late 2007 cooling:

And here’s the LW (infrared) component…note the very low emission late in 2007, a portion of which must be from the colder atmosphere emitting less infrared radiation.

As I discuss at length in the paper I am preparing, the physical interpretation of which of these 3 processes is dominant is helped by drawing a phase space diagram of the Net (LW+SW) radiative flux anomalies versus temperature anomalies (now shown as monthly running 3-month averages), which shows that the 2007-08 cooling event has a classic radiative forcing signature:

The spiral (or loop) pattern is the result of the fact that the temperature response of the ocean lags the forcing. This is in contrast to feedback, a process for which there is no time lag. The dashed line represents the feedback I believe to be operating in the climate system on these interannual (year-to-year) time scales, around 6 W m^-2 K^-1 as we published in 2007…and as Lindzen and Choi (2009) recently published from the older Earth Radiation Budget Satellite data.

The ability to separate forcing from feedback is crucial in the global warming debate. While this signature of internal radiative forcing of the 2007-08 event is clear, it is not possible to determine the feedback in response to that temperature change – it’s signature is overwhelmed by the radiative forcing.

Since the fluctuations in Net (LW+SW) radiative flux are a combination of forcing and feedback, we can use the tropospheric temperature variations to remove an estimate of the feedback component in order to isolate the forcing. [While experts will questions this step, it is entirely consistent with the procedures of Forster and Gregory (2006 J. Climate) and Forster and Taylor (2006 J. of Climate), who subtracted known radiative forcings from the total flux to isolate the feedback].

The method is simple: The forcing equals the Net flux minus the feedback parameter (6 W m^-2 K^-1) times the LT temperature variations shown in the first figure above. The result looks like this:

What we see are 3 major peaks in radiant energy loss forcing the system: in 2000, 2004, and late 2007. If you look at the features in the separate SW and LW plots above, it is obvious the main signature is in the SW…probably due to natural increases in cloud cover, mostly low clouds, causing internal radiative forcing of the system

If we instead assume a much smaller feedback parameter, say in the mid-range of what the IPCC models exhibit, 1.5 W m^-2 K^-1, then the estimate of the radiative forcing looks like this:

Note the trend lines in either case show a net increase of at least 1 W m^-2 in the radiant energy entering the climate system. The anthropogenic greenhouse gas component of this would be (I believe) about 0.4 W m^-2, or a little less that half. I’ll update this if someone gives me a better estimate.

So, what might all of this mean in the climate debate? First, nature can cause some pretty substantial forcings…what if these occur on the time scales associated with global warming (decades to centuries)?

But what is really curious is that the 9-year change in radiative forcing (warming influence) of the system seen in the last two figures is at least TWICE that expected from the carbon dioxide component alone, and yet essentially no warming has occurred over that period (see first illustration above). How could this be, if the climate system is as sensitive as the IPCC claims it to be?

YR MON GLOBE NH SH TROPICS
2009 1 +0.304 +0.443 +0.165 -0.036
2009 2 +0.347 +0.678 +0.016 +0.051
2009 3 +0.206 +0.310 +0.103 -0.149
2009 4 +0.090 +0.124 +0.056 -0.014
2009 5 +0.045 +0.046 +0.044 -0.166
2009 6 +0.003 +0.031 -0.025 -0.003
2009 7 +0.412 +0.212 +0.610 +0.427
2009 8 +0.231 +0.284 +0.179 +0.455

August 2009 saw a modest fall in the global average tropospheric temperature anomaly, from +0.41 deg. C in July to +0.23 deg. C in August. The tropical and Northern Hemispheric troposphere remain quite warm, but the Southern Hemisphere cooled by over 0.4 deg. C in the last month.

NOTE: For those who are monitoring the daily progress of global-average temperatures here, we are still working on switching from NOAA-15 to Aqua AMSU, which will provide more accurate tracking on a daily basis. We will be including both our lower troposphere (LT) and mid-tropospheric (MT) pre-processing of the data. We will also be adding global sea surface temperature anomalies from the AMSR-E instrument on board the NASA Aqua satellite.

This is the first of what might turn into a series of monthly updates of some maritime climate parameters monitored by the AMSR-E instrument on NASA’s Aqua satellite. All monthly statistics have been computed by me from daily global gridpoint data produced and archived by Remote Sensing Systems (RSS) under the direction of Frank Wentz, a member of our U.S. AMSR-E Science Team. Since Aqua was launched in 2002, the data are available only since June, 2002. A description of how these products were derived, and where they reside, is provided here.

There are 5 “ocean products”: sea surface temperature [SST]; near-surface wind speed; vertically-integrated water vapor; vertically integrated cloud water; and rain rate. I will present time series of monthly anomalies averaged over the global, ice-free oceans (56 deg. N to 56 deg. S latitude), and separately for the deep tropics (20 deg. N to 20 deg. S latitude). ‘Anomalies’ are departures from the average seasonal cycles in those parameters, which will be recomputed as each new month of data is added.

GLOBAL OCEANS

In the first figure below are plotted the 5 ocean products for the global ice free-oceans (56N to 56S). As can be seen in the top panel, SSTs in August cooled slightly from the unusually warm conditions experienced in July.

I have added linear trend lines to each time series, which you are free to misinterpret as you wish. 😉 Since the AMSR-E period of record is only 7.25 years long, a calculated trend won’t have much meaning…although it will be interesting to see how long it takes before the climate system obeys the UN’s command to warm, and the SST trend line begins to go uphill again.

How these different variables change relative to each other is illustrated in the following lag-correlation plot of SST versus the other variables. “PDO” is the Pacific Decadal Oscillation Index, while “SOI” is the Southern Oscillation Index (negative for El Nino, positive for La Nina). A discussion of these curves is provided later, below.

TROPICAL OCEANS

The next figure shows the ocean product anomalies for just the deep tropics, 20N to 20S latitude….

…and the lag correlation plot for the deep tropics is next:

DISCUSSION

Using the 20N-20S lag correlation plot as an example, you can see that total integrated water vapor is highly correlated with SST, which in turn is highly correlated with El Nino conditions (negative SOI values).

Also note that sea surface temperature tends to peak after months of anomalously low wind conditions, then falls as wind speeds increase.

Cloud water and rain rates increase as SST increases, reaching a maximum 1 to 3 months after the SST peaked.