Northwest Climate Change: Did the 2014 National Climate Assessment Get the Story Right?

This week the media has been full of stories about the release of the 2014 National  Climate Assessment, which examines recent and future climate change around the nation.  At 841 pages, this is a large body of work and in this blog I will examine the portion dealing with the Pacific Northwest.  Did it get the facts correct? Was its handling of the material unbiased, hyped, or too conservative?   Let's take a look.

This section begins with a review of climate change over the past century.  Specifically, they note:

Temperatures increased across the region from 1895 to 2011, with a regionally averaged warming of about 1.3°F.   While precipitation has generally increased, trends are small as compared to natural variability ... Studies of observed changes in extreme precipitation use different time periods and definitions of “extreme,” but none find statistically significant changes in the Northwest.

Fair enough.  Over a more than a century Northwest-averaged temperature has only warmed by about 1.3F and there have been no significant trends in precipitation, including extreme precipitation.  And they note that not all of the 1.3F rise in temperature was due to human influences;  some of that warming was undoubtedly natural since it started in the late 1800's before humans were having much impact.  To illustrate this, the figure below (from the UW Climate Impacts Group) demonstrates that much of the warming occurred before 1940, prior to significant effects of human emissions of greenhouse gases.

So Northwest climate has really not changed much during the past century.

Then the report talks about future projections:

An increase in average annual temperature of 3.3°F to 9.7°F is projected by 2070 to 2099 (compared to the period 1970 to 1999), depending largely on total global emissions of heat-trapping gases. 

So they are suggesting that temperatures over that 100 year period (basically the difference of 1985 and 2085) will warm at 2.5 to 7.5 times the rate it did over the 1895 to 2011 period.  They do not give references for this warming, but we can examine the the co-authors' (Phil Mote and Amy Snover) previous work, which give similar numbers.

 Below is a figure from the recent UW Climate Impact's Group regional climate evaluation (Any Snover is the head of the CIG).  It shows Northwest temperatures predicted by global climate models for various scenarios (increases in greenhouse gases).  8.5 is the most extreme, 2.6 the least.  8.5 is the scenario where we don't make drastic changes in our life style and I believe the most probable assumption. The range of climate model solutions for the 8.5 and 4.5 scenarios are shown and are consistent with the recently released climate assessment report.

You will immediately note a problem with their plot, the temperatures over the NW have been flat since 1985 (see their figure above) and the figure below show a roughly 2F increase.  Furthermore, I believe it is doubtful that the warming will be as much as shown by the 8.5 simulation (7-11F warming by 2085) for two reasons.  First, the climate models appear to be too sensitive to greenhouse gas forcing and secondly they appear to have particular trouble with natural variability of the eastern Pacific.

A number of climate experts at the UW (such as Professor K.K. Tung, paper here) believe that excessive sensitivity of climate models to greenhouse gases results from their "tuning" to match the  warming of the late 20th century.  They believe that the sensitivity of the models was cranked up to compensate for either the failure to handle the cooling effects of atmospheric particles (aerosols) or the models' inability to properly simulate major modes of natural variability like the Atlantic Multidecadal Oscillation (AMO) mode. The tuned models look great over the 20th century (proof you can trust them!), but they overdo things in the future.

In addition, the climate models runs that included the past 40 years have failed to produced the observed cooling of the eastern Pacific (see figure), which obviously is suppressing temperatures around here.

I suspect cutting the RCP 8.5 warming prediction in half would be a far more reasonable prediction of what things will be like in 2085 (as well as starting at the observed level of 2014).  So we are talking 2-4F warming.  The problem is the authors of this section use the larger warming estimates as assumptions in much of the rest of their work (like snowpack changes).

The assessment notes that expected precipitation changes in the region will be modest, with a small increase in precipitation being probable:

Change in annual average precipitation in the Northwest is projected to be within a range of an 11% decrease to a 12% increase for 2030 to 2059 and a 10% decrease to an 18% increase for 2070 to 
2099 for the B1, A1B, and A2 scenarios.

This is an area where the assessment is probably too conservative.  Although annual precipitation will not change much, some research (including that of my group), suggests substantial enhancement of the heaviest precipitation events (pineapple express).  Furthermore, with less snow pack on the slopes during early winter, enhanced rainfall could result in more serious flooding (snow actually soaks up some rain, reducing flood risk).

In the next section, the assessment states that the snowpack has declined about 20% since 1950.  I co-authored a peer reviewed paper on the subject and found a reduction of 23% since 1930.  But what is not said in the Climate Assessment Report is that most of this 23% has nothing to do with global warming, but rather was forced by natural variability.  They claim that spring snowmelt has occurred 0 to 30 days earlier since 1950 depending on location, while we found snow melt occurs only 5 days early since 1930. Strangely, they don't even cite our paper in their section, which suggests either a biased viewpoint or a very incomplete search of the literature, neither of which is good.

 The report talks about increases in Northwest wildfires:

Although wildfires are a natural part of most Northwest forest ecosystems, warmer and drier conditions have helped increase the number and extent of wildfires in western U.S. forests since the 1970s

Reading through the wildfire literature the last few days and talking to experts in the USDA Forest Service, I could find no research that shows a significant increase in wildfires in the Northwest over the three decades.   And even if they did increase, it would be nearly impossible to separate climate impacts from the influence of increasing population, changes in forestry practices, and major shifts in wildfire management.  Wildfires are a natural aspect of NW ecology, and the suppression of fires has resulted in the potential for catastrophic conflagrations.  Quite possibly, more fires would be a good thing for the environment.

But most significantly, this regional climate report fails to mention the central and most important fact about climate change in the Northwest:  the impacts of global warming will be weaker and more delayed in the Northwest than nearly any location in the U.S.   The reason:  the Pacific Ocean.

Virtually all climate models indicate the arctic will warm up the most and the continents will warm up more than the oceans (see figure).  Furthermore, the eastern oceans will warm more slowly than the western

oceans.  The weather of the Pacific Northwest is controlled by the Pacific Ocean to our west, and if the Pacific warms more slowly so will we.  In fact, the lack of east Pacific warming of the past 30 years explains why our temperatures and snow pack has remained essentially constant during that period.  Furthermore, most Northwest climate change of the last century has been natural in origin.

The bottom line of my analysis is that the impacts of increased greenhouse gases has been very small here in the Northwest so far.  Global warming will impact our region, but the changes here will be modest and slow. The biggest impacts will be reduced snow pack later in the century and increased risk of heavier precipitation and floods.   Less snow pack will reduce the amount of snow melt available during the summer and early fall. Thus, to ensure sufficient water supplies we will need larger or additional reservoirs or dams.