The Great Russian Heat Wave of July 2010
The July 2010 heat wave for western Russia, including Moscow, is one of the most impressive heat waves I've ever seen. I thought it would be instructive to look at the meterological conditions to understand what creates heat waves and droughts.
Fig. 1 Meteorological history of July 2010 for Sheremetyevo International Airport, Moscow, Russia.
Theory of Drought
The most immediate cause for drought/heat waves are upper-level high pressure systems, colloquially known as "ridges". Since winds move from high to low pressure, this means winds are moving away from upper-level winds. In response, air starts descending beneath the ridge, this is known as subsidence In the free atmosphere, as a parcel of air descends, it starts to warm, decreasing the relative humidity, and making it more difficult for clouds to form. The apocryphal quote from Bill Gray about this is, "Up moist, down dry". So with large amounts of subsidence under the ridge, there aren't as many clouds to shield the ground from the sun. So the ground warms up, which warms the air.
Now positive feedback loops start to form. The first is that increasing the air temperature strengthens the upper-level ridge, which warms the ground more. Another loop is that with no clouds, there is no rain, which causes the ground to dry out. As the ground dries out, it warms more quickly for the same amount of sunlight. Also as the ground dries out, moisture near the surface drops, which makes precipitation less likely and further dries out the surface continuing the heat wave/drought.
Short of waiting for the seasons to change, the best way to get rid of a heat wave is for a change in the upper-level circulation to help push the ridge away from the drought circulation. If the heat wave is relatively young (less than 2 weeks old), a passing tropical cyclone can provide enough moisture to wet the ground and break the soil moisture feedback loop.
Before we can know what's not normal, we have to know what's normal first.
To calculate the normals, 30-year climatologies of different meterological quantities (i.e. soil moisture and temperature, surface temperature, and 500 mb heights) were calculated using data from the Climate Forecast System Reanalysis (CFSR) (NOMADS data repository). Here is a description how a reanalysis works. CFSR is notable because it is the first reanalysis to use a coupled atmosphere-ocean model. As a result, CFSR has physically consistent estimates of the conditions of the atmosphere, ocean, and land. CFSR has data from 1979 (When polar-orbiting satellites became able to estimate vertical profiles of temperature) to 2009 (NCEP expects to start releasing 2010 data soon.)
Data for June and July 2010 were taken from the GFS Data Assimilation System (GDAS). GDAS combines observations with short-term forecasts (3-6 hours) from the GFS to provide initial conditions for the next model run of the GFS (i.e., the 6Z run makes the 12Z initial conditions). Deviations from normal were computed for each day for different quantities and then averaged by month.
In comparing deviations from normal across wide regions, it helps to normalize the deviations. A temperature deviation of 3 degrees C may be not that unusual in one region, but may be very significant in another. The solution is to use climatological anomalies. Calculating the climatological anomaly is a two step process. First, we calculate the difference between a quantity (i.e., temperature) and it's 30-year average value. Then we normalize the difference by dividing it with the 30-year standard deviation. From statistical theory, we know how unusual climatological anomalies are by value:
Odds of a deviation > 1 climatological anomaly=31.7%
Odds of a deviation > 2 climatological anomalies=4.5%
Odds of a deviation > 3 climatological anomalies=0.27%
Odds of a deviation > 4 climatological anomalies=6.34/1000%
Odds of a deviation > 5 climatological anomalies=5.7/100000%
Odds of a deviation > 6 climatological anomalies=1.9/1000000%
In Fig. 2, we can see a large area of +1 anonmalies in average 2-m temperature (the temperature recorded in weather reports) across western Russia, with small areas of +2 anomalies around Moscow. This means that for Moscow, the chances of the monthly temperature being this warm is less than 4.5%.
Fig. 2 Climatological anomalies of average 2-m temperature for July 2010.
Looking at Fig. 3, we see that the the 500 mb heights are greater than expected in the heat wave region. This shows that a stronger than usual upper-level ridge is in place over the region, initiating the heat wave. The low height anomalies to the west and east of the ridge indicate that an Omega block was in place for much of July, keeping the ridge in place. Also, looking at the pattern of the height anomalies, it is clear that the general circulation of the northern mid-latitudes is different than the typical summer patterns.
Fig. 3 Average deviations (not normalized) of 500 mb heights for July 2010.
Figure 4 shows that the soil moisture feedback loop was active for the month of July. There are widespread anomalies of -3 and even small areas of -4. This means that 99.8% of the time, the soil is more moist than it is now. This is an exceptionally unusual soil moisture shortfall, and it's occurring in in Russia's wheat growing region.
Fig. 4 Climatological anomalies of soil moisture for July 2010.
Now the keen-eyed reader may be saying, "This is all fine and good model analysis, but do you have any real data?" The MODIS instrument on NASA's TERRA and AQUA can measure how green vegetation is from space. Figure 5 shows that the vegetation in southern Russia is much browner than usual for the month of July. This is consistent with the temperature and soil moisture anomalies described earlier.
Fig. 5 NDVI anomalies for July 2010 from MODIS data over Southern Russia (provided by NASA's Earth Observatory)
Global Warming and Heat Waves
As the climate warms, we expect heat waves to become more frequent (Ganguly et al., 2009). Now there is still considerable uncertainty on where the heat waves will occur, that seems to depend on the climate model used. However, the physics of heat waves do not change. Heat waves in climate simulations are still associated with upper-level ridges (Meehl and Tebauldi, 2004). This suggests that we will likely see more heat waves like the Muscovite heat wave of 2010 in the future.
Hong, Song-You, Eugenia Kalnay, 2002: The 1998 Oklahoma–Texas Drought: Mechanistic Experiments with NCEP Global and Regional Models. J. Climate, 15, 945-963
Ganguly et al., Higher Trends But Larger Uncertainty And Geographic Variability In 21st Century Temperature And Heat Waves, Proceedings of the National Academy of Sciences, 2009.
G. Meehl and C. Tebauldi, More Intense, More Frequent, And Longer Lasting Heat Waves In The 21st Century, Science, 2004.
J. Namias, "Some Meteorological Aspects Of Drought With Special Reference to The Summers Of 1952–54 Over The United States", Monthly Weather Review, 1955.
J. Namias, "Anatomy of Great Plains Protracted Heat Waves (Especially The 1980 U.S. Summer Drought)", Monthly Weather Review, 1982.
Palecki, Michael A., Stanley A. Changnon, Kenneth E. Kunkel, 2001: The Nature and Impacts of the July 1999 Heat Wave in the Midwestern United States: Learning from the Lessons of 1995. Bull. Amer. Meteor. Soc., 82, 1353-1367
W. Palmer, "Meterological Drought", US Weather Bureau Research Paper No. 45, 1965