Downslope Winds and High Pressure Domes

Air Compression: The Neglected Climate Heat Source

Roy Clark

A basic misconception that has been widely promoted in both the climate literature and the news media is that an increase in the atmospheric concentration of CO₂ will lead to an increase in the intensity and frequency of ‘extreme weather events’ [Herring et al, 2022]. At present the average annual increase in CO₂ concentration is approximately 2.4 parts per million (ppm) per year. This produces an increase in the downward LWIR flux to the surface near 0.034 W m^-2 per year. Such a small increase in flux can have no measurable effect on any ‘extreme’ weather system. Good examples of these false ‘extreme weather’ claims are increases in brush fires and heat waves. These claims conveniently overlook the main source of heat that drives these events: that produced by air compression during downward flow in the lower troposphere.

When dry air ascends through the troposphere by convection, the lapse rate or cooling rate is 9.8 °C km^-1. For each kilometer increase in altitude, the temperature decreases by almost 10 °C. The same is true in reverse. For a decrease in altitude of 1 km, dry air warms by almost 10 °C. The temperature changes are produced by air expansion or compression. As the air rises and expands it must perform mechanical work. The energy required is removed from the motion of the air molecules, both the kinetic energy and the internal vibration/rotation. This is converted into gravitational potential energy. The reverse occurs when the air descends and is compressed. There are two processes that produce a downward flow of warm air to the surface. The first is a downslope wind that flows down the side of a hill or mountain. The second is the circulation of air within a high pressure circulation system or ‘dome’. In this case, Coriolis forces related to the air rotation produce the downward flow. Both of these processes are discussed in ‘Finding Simplicity’. Section 5.4 examines the onshore/offshore flow transition in Southern California. Section 5.5 considers the effect of a high pressure dome on temperatures measured at the Woomera weather station, S. Australia. These sections are posted below.

Downslope winds are well known in many regions or the world and there are different names for the same effect. In S. California they are Santa Ana Winds. In N. California they are diablo winds. In the Rocky Mountains they are chinook (‘snow eating’) winds. In the Alps they are föhn winds. A good example of the effect of downslope winds on temperature was recorded at Havre, Montana, December 16 to 18, 1933 [Math, 1933]. At this time the CO₂ concentration was near 310 ppm. The thermograph trace is shown in Figure 1a. The temperature first rose by 27 °F in five minutes and increased by a total of 53 °F in less than 2 days. The temperature then cooled by 41 °F in two hours. There is no connection between these downslope wind events and any increase in atmospheric CO₂ concentration. Once the necessary weather pattern is established, the hot, dry winds will dry out the vegetation very quickly and any ignition source will start the fire. In S. California, a high pressure system over the Great Basin produces an offshore flow that descends from the desert plateau. The winds may be increased by an adjacent low pressure region. This is illustrated in Figure 1b. Figure 1c shows a Terra Satellite image taken 12/5/17 showing the fires in S. California [Thomas Fire, 2017]. The smoke is blown out to sea by the offshore winds. The Marshall fire in Boulder Colorado, December 30, 2021 that destroyed about 1000 houses was caused by strong downslope winds and an ignition source related to human activity. The fuel was dry grass and any residual moisture would have been removed very quickly by the dry 100 mph winds [Mass, 2022].

A high pressure dome formed over the Pacific Northwest in late June 2021. This produced record high temperatures as shown in Figure 2 [Watts, 2021]. As the high pressure system moved east, the temperature in Portland OR dropped from 116 to 64 °F over the night of June 28 to 29. Once a ‘blocking’ high pressure system pattern is established, it can persist for weeks or even months. Since these systems also block rainfall and remove soil moisture, additional heating is produced by the reduced latent heat flux at the surface. For example, there was nothing unusual about the 2003 European heat wave [Black et al, 2004]. Brush fires produced by ‘blocking’ high pressure systems are a normal part of the Australian climate [Foley, 1947]. Similarly, a high pressure system regularly forms over the area near Verkhoyansk, Siberia. This produces very high summer temperatures and very low winter temperatures [Autio, 2020 Watts, 2020].

References

Autio, P. (2020), “Siberia on fire – every summer” Watts Up With That Post July 14, [https://wattsupwiththat.com/2020/07/14/siberia-on-fire-every-summer/] Autio

Black, E., M. Blackburn, G. Harrison, B. Hoskins and J. Methven (2004), "Factors contributing to the summer 2003 European heatwave". Weather 59(8), pp. 217-223. [https://doi.org/10.1256/wea.74.04] Black

Foley, J. C. (1947), “A study of meteorological conditions associated with bush and grass fires and fire protection strategy in Australia” BOM Bulletin 38, [https://nla.gov.au/nla.obj-257165724/view?partId=nla.obj-257176204#page/n4/mode/1up] Foley

Herring, S. C., N. Christidis, A. Hoell and P. A. Stott (2022), “Explaining Extreme Events of 2020 from a Climate Perspective” Bull. Amer. Meteor. Soc. 101 (1), pp. S1–S128. [https://doi.org/10.1175/BAMS-ExplainingExtremeEvents2020.1] (and prior years in this series) Herring

Mass, C. (2022), “The Colorado Wildfire and Global Warming: Is there a Connection?” Watts Up With That, Post Jan. 6. [https://wattsupwiththat.com/2022/01/06/the-colorado-wildfire-and-global-warming-is-there-a-connection/] Mass

Math, F. A. (1934), “Battle of the chinook wind at Havre, Mont.” Monthly Weather Review Feb. 1934 pp. 54-57, [https://doi.org/10.1175/1520-0493(1934)62<54:BOTCWA>2.0.COsemicolon2] (semicolons do not work with this text editor)

Thomas Fire (2017), [https://en.wikipedia.org/wiki/December_2017_Southern_California_wildfires] ThomasFire

Watts, A. (2021), Watts Up With That, Post Jun. 30. [https://wattsupwiththat.com/2021/06/30/major-media-fail-on-reporting-the-pacific-northwest-heatwave/] Watts.2021

Watts, A. (2020), “Climate Change? Temperature Hits 100 Degrees above Arctic Circle, Just Like 100 Years Ago” Watts Up With That Post, June 23. [https://wattsupwiththat.com/2020/06/23/climate-change-temperature-hits-100-degrees-above-arctic-circle-just-like-100-years-ago/] Watts.2020

Excerpts from 'Finding simplicity in a Complex World'

5.4 The Onshore/Offshore Flow Transition in Southern California

The Desert Rock Cases illustrate the seasonal variation of the DES attractor and show that the temperature drift is determined by the local weather patterns. The Grasslands data set is a full year of temperature and flux data recorded at an advanced AmeriFlux monitoring site located in Limestone Canyon Regional Park, near Irvine, S. California in 2008 [Clark, 2013]. The complete data set consisted of half hour averages of 17 parameters: friction velocity air temperature wind direction wind speed CO₂ flux H₂O flux sensible heat flux latent heat flux CO₂ concentration H₂O concentration incoming photosynthetic active radiation reflected photosynthetic active radiation incoming global solar radiation reflected global solar radiation relative humidity precipitation and net radiation. The maximum and minimum daily air temperatures and the 8 day satellite average min/max skin (surface) temperatures for the full data set are shown in Figure 5.37. The various temperature spikes through the year are caused by the transition from coastal onshore winds to desert offshore winds with a decrease in humidity and an increase in air temperatures from downslope wind effects. The heating is caused by air compression as the dry air descends ~1 km from the desert to the coast. The minimum surface and air temperatures are similar, but the maximum surface temperature is higher than the maximum air temperature. The surface temperature also increases with the air temperature, although the surface temperatures are 8 day averages that do not change as fast as the air temperatures.

5.4.1 The net LWIR and the Latent Heat Fluxes

Changes in humidity and cloud cover alter the net LWIR cooling flux emitted by the surface. The magnitude of the net LWIR flux increases with decreasing humidity and decreases with increasing humidity and cloud cover. The average daily night time net LWIR fluxes for 2008 are shown in Figure 5.38. The annual average is 44 ±16 W m^-2 (1σ standard deviation). The high values for the net LWIR flux are characteristic of offshore flow conditions associated with strong downslope winds known as Santa Ana winds. The low values indicate cloud cover or a marine layer over the measurement site.

Most rainfall in S. California occurs during the winter months. The vegetation then dries out over the summer leading to increased fire risk during offshore flow conditions. The cumulative daily latent heat fluxes for daytime and nighttime conditions are shown in Figure 5.39a. The ratio of the latent heat flux to the solar flux is shown in Figure 5.39b. Almost all of the evaporation occurs during the day. The peak latent heat flux occurs in March as the soil and vegetation dry out after the winter rains. The maximum value of the latent heat flux is 6 MJ m^-2 day^-1 and maximum ratio to the solar flux is 0.4.

5.4.2 The Onshore - Offshore Transition during Days 76 to 85

The 30 minute air and estimated surface temperature data for days 76 to 85, 2008 from data recorded at the Grasslands AmeriFlux site are shown in Figure 5.40. The surface temperatures were not recorded and are estimated from the air temperatures and the increase in daytime net IR emission from the surface. There was a 10 °C increase in minimum air temperature from day 81 to day 83. This was the period of transition from onshore to offshore flow conditions. The convection transition temperature increased by 10 °C from 10 to 20 °C. Based on the analysis of the Desert Rock data, Figures 5.18b and d, the minimum air temperature is generally a good indicator of the minimum surface temperature.

The flux terms, RH and wind speed data for days 76 to 85 are shown in Figure 5.41 and the cumulative daily flux terms are shown in Figure 5.42. Figure 5.41a shows the measured solar flux and the calculated values from Eqn. (4.4). There is good agreement between the measured and calculated values. Figure 5.41b shows the IR flux. The flux profile is uneven and the daytime values do not track the solar flux. There are several factors that may contribute to this. First the site is on a west facing hillside slope with higher hills to the east. The surface is also a mix of rough grass, chaparral and exposed areas of soil and rock. The solar illumination of the surface is therefore uneven. This requires further study. The decrease in IR flux for the start of day 80 indicates higher humidity/cloud cover. Figure 5.41c shows the sensible heat (SH) and latent heat (LH) fluxes. The combined SH+LH flux is shown in Figure 5.41d. There is a decrease in the SH flux associated with the offshore conditions on days 81 through 84. This can be seen more clearly in the cumulative daily flux data shown in Figure 5.42, yellow line. The relative humidity from Figure 5.41e shows a peak for the night at the start of day 80. This is consistent with the decrease in net IR flux. The wind speed from Figure 5.41f is generally low at night and increases during the afternoons. The land surface warms through solar heating and convection establishes an onshore flow during the afternoon. There is an overall increase in wind speed associated with the offshore flow. Inspection of the wind direction (not shown) for days 81 to 83 indicated that there was an offshore flow that reversed during the afternoon. The data are based on single point measurements with instruments mounted on a 10 m tower. The total SH+LH+net IR flux is less than the net solar flux. Additional instrumentation is needed to properly characterize the complex energy transfer processes that occur at this site and achieve closure for the flux balance.

5.4.3 A Simple Model of the Onshore-Offshore Temperature Transition

We may use our simple single reservoir thermal model to investigate the onshore offshore flow transition. Here we use the measured net IR flux directly as a model input. We calculate the surface temperature by using the average night time net IR flux to estimate the LWIR window transmission. We then combine this with the daytime air temperature and the increase in daytime net IR flux to estimate the daytime surface LWIR flux and from this, we estimate the temperature.

T_surf = [((σT_air⁴ + net IR - IR_nightav)/0.95)/σ]^¼ .............................................................Eqn. (5.1)

This was used to calculate the surface temperature shown in Figure 5.40. The SH flux was calculated from Eqn. (4.8). The SH flux algorithm was constrained so that it was only used in the model if the value was positive and the solar flux was above a threshold of 50 W m^-2. The best least squares fit to the surface temperature was obtained with k_evp set to 35 W m^-2 C^-1 and a reservoir heat capacity of 0.29 MJ m^-2. The results are shown in Figure 5.43. The correlation coefficient was 0.48.

The calculated surface temperature shows the offshore to onshore flow temperature transition. However, the use of the single thermal reservoir delays the thermal response because of the higher heat capacity. The surface layer responds faster in the measured data.

References

Clark, R. (2013), “A dynamic, coupled thermal reservoir approach to atmospheric energy transfer Part II: Applications” Energy and Environment 24(3, 4) pp. 341-359. [https://doi.org/10.1260/0958-305X.24.3-4.341] Clark

5.5 The Effects of Blocking High Pressure Systems on the Surface Temperature for Woomera, S. Australia

The Grasslands data shows the rapid warming effect of downslope winds on the surface temperature. Here we investigate the warming produced by the downward air flow within a blocking high pressure system using data from Woomera, S. Australia for 2018 and 2019. The formation of a blocking high pressure system over the Australian Bight is a common occurrence. These systems can produce elevated temperatures over a wide area of Australia and lead to increased risk of brushfires [Fox-Hughes, 2014]. Here we consider the effect on surface temperature of two such blocking high pressure systems that were recorded at Woomera, S. Australia in December 2018 and 2019. The 2019 event was also associated with a record high value near 2 °C for the Indian Ocean Dipole (IOD) Index. This meant that the western Indian Ocean was warmer than the eastern part. This reduced ocean evaporation near Australia and contributed to drought conditions over parts of Australia. The IOD Index from the start of 2018 to June 2020 is shown in Figure 5.44.

The daily weather station data for Woomera, Station # 16001, for 2018 and 2019 were downloaded from the Australian Bureau of Meteorology website. The data consisted of minimum and maximum temperatures, precipitation and total daily solar flux. The minimum and maximum temperatures for 2018 and 2019 are shown in Figure 5.45. Peak temperatures occur in December and January (seasons are reversed in the S. Hemisphere). The temperatures also show short term variations of approximately 10 °C that are caused by downslope winds and/or blocking high conditions. The total daily solar flux and precipitation for 2018 and 2019 are shown in Figure 5.46. The envelope of the solar flux indicates the flux for cloud free conditions and the decreases are produced by clouds. Rainfall is low and can occur at any time of the year.

The data for December 2018 and 2019 from Figures 5.45 and 5.46 are shown in Figure 5.47. Blocking high warming conditions occurred in 2018 for days 356 to 362 with a peak temperature of 46.2 °C on day 362. Similar conditions occurred in 2019 for days 346 to 354 with a peak temperature of 48.2 °C on day 354. The blocking high events are indicated by the dotted lines and the arrows. The 11 day temperature profiles of the two blocking high events are shown overlapped in Figure 5.48. The 2019 event lasted 2 days longer than the one in 2018, based on the rise in the maximum temperature. The downward air flow and compression within the blocking high pressure dome for the extra two days led to record high temperatures. Drought conditions associated with the large positive value for the IOD index in 2019 also contributed to the high temperatures (see Figure 5.44).

In order to simulate the data sets from Figure 5.48 we used the solar flux data for days 162 to 172 (2019) and 172 to 182 (2018) at 35° latitude calculated from Eqn. (4.4). The evapotranspiration was calculated using Eqn. (4.10) and the model was also configured so that the T_d values could be changed at each day at midnight to simulate the daily change in convection transition temperature. The daily minimum temperatures were used for the T_d values based on the results from the Desert Rock data, Figures 5.18b and 5.18d. These show that the minimum air and surface temperatures are usually similar. The evapotranspiration flux values were constrained so that negative values were set to zero. The model was adjusted using a least squares fit to the measured minimum temperatures. The best fit to the data was obtained with k_evt set to 320 W m^-2 for both the 2018 and 2019 data sets. The calculated surface temperatures over the 11 days and the measured minimum and maximum surface air temperatures are shown in Figure 5.49. The measured and calculated maximum and minimum temperatures are summarized in Figure 5.50. Our simple single thermal reservoir with T_d values based on measured minimum air temperatures simulates the expected features of the temperature profile. The calculated T_min values are near the measured values and the maximum surface temperatures are larger than the maximum values of the measured air temperature. This is similar to the Desert Rock data for Case 2.