Climate Monitoring

One of the exciting aspects of modern climate science is that the tools for monitoring global climate have advanced by leaps and bounds in recent years.  Regular readers know that I'm a fan of the ECMWF's ERA5 reanalysis, which is a global gridded model estimate of hourly weather across the globe, and extending back to 1950.  The ERA5 product even includes uncertainty information based on an ensemble of possibilities that are consistent with the observed historical data.  ERA5 is certainly not without flaws, some of which are obviously connected to the 31-km grid spacing, but for many purposes it is a tremendous tool.

To illustrate what ERA5 can deliver, here's a look at June climate anomalies for a few variables that we don't usually monitor in detail, because they are not widely measured at ground level with consistent and reliable instrumentation.  For example, solar radiation:

The map shows the percentile rank of the June total solar (shortwave) radiation compared to the 1991-2020 distribution for the month of June.  On this scale, 0% would mean that June saw less solar radiation than any June in the past 30 years, and 100% would indicate it was the sunniest June in more than 30 years.  So according to ERA5, June was a very cloudy month along the west coast and western interior, as well as southwestern and south-central Alaska.

ERA5 precipitation shows a broadly similar pattern, with very wet conditions across the eastern Bering Sea, northwestern Alaska, and also in southeast Alaska.

Interestingly the NOAA/NCEI climate division data for June (see below) shows some significant differences, including much drier conditions from the Aleutians to south-central Alaska, but less dryness in the eastern interior.  Looking at some ground truth data, Anchorage was certainly dry with only 0.31" of rain (5th driest relative to the 30 years), and Homer also saw less rain than normal.  However, King Salmon was only a bit drier than normal, Kodiak was wetter than normal, and Cold Bay rainfall was substantially above normal (4th wettest).

We can also look at ERA5 wind speed, bearing in mind that presumably only the large-scale wind anomalies have much validity because of the model's inability to represent complex terrain.

 

Here's dewpoint: above normal almost everywhere as a consequence of moist air being drawn up from the south.

The map below shows the mid-atmosphere pressure anomaly that produced the humid air flow and the wet, cloudy weather in the west and northwest: an unusually strong trough near the Bering Strait, and a strong ridge over southwestern Canada.  There was also very unusual low pressure in the Arctic, as highlighted in my comment a couple of weeks ago about the wet weather in the northwest.

Here's an example of an ERA5 variable that gives another perspective on the June climate anomalies:

According to the model, evaporation was well below normal in the southwest, but it was not widely suppressed elsewhere despite cloudy weather being quite widespread, and evaporation was higher than normal in the southeastern interior even though the dewpoint was high (because it was also warm - see figures at bottom).  But the far northern Yukon is a bit of a puzzle, as ERA5 shows below-normal evaporation despite low humidity and above-normal sunshine, wind, and temperature.  The surprisingly low evaporation may be related to low soil moisture, i.e. the the model thinks the ground was so dry that evaporation was reduced for lack of soil moisture.  The maps below show May and June soil moisture in the top layer of the ERA5 land model.

Finally, temperature - see below.  The ERA5 result compares reasonably well to the climate division data, except (and it's a big exception) on the North Slope.  Umiat was definitely warmer than normal, so this looks like a topic for further investigation.

Update: Rick Thoman's station plot for June showed near-universal warmth on the North Slope.  So I'm not sure what's going on with the climate division data, but I hope to find out.