Cold-Air Pools in Mountain Valleys

On clear nights with calm winds, the ground cools rapidly. Air in contact with the colder ground cools by conducting heat to the ground. This process can continue until sunrise.

When this cooling process occurs along mountain slopes, the cooling air becomes colder and denser than the air away from the slopes, which causes the cold air to sink downslope. The dense cold air flows downslope in streams (called katabatic winds) following the steepest slopes. When the cold air flows into a relatively flat area (a mountain or river valley, for example), the streams of cold air slow down. This causes the valley to fill with cold air, much like streams filling a lake.

Mountain valleys around the world often fill with cold air a few hundred to thousands of feet deep, depending on the depth of the valley and the shape of the valley allowing cold air to flow out of the valley. Above this “cold air pool,” the air remains warmer because is not in contact with the ground. This means that low elevations in mountain valleys are regularly exposed to cold temperatures at night and higher elevations above the cold air pool remain in warmer air when weather conditions are clear and calm.

Meteorological observations of cold air pool formation and evolution were taken in the mountain valley of Hubbard Brook Experimental Forest, New Hampshire in November 2015. We deployed a tethered weather balloon to 150 meters (~500 feet) with temperature sensors tied to the nylon string every 5 meters (~16 feet; Fig. 1) to record the thermal evolution and depth of the cold pool overnight and its dissipation the next morning. We also measured wind, temperature and relative humidity on the north and south facing slopes surrounding the valley. We worked in pairs all night taking weather observations on the mountain slopes and measuring the balloon elevation angle – an indication of wind at the balloon’s altitude.

Figure 1:. Hubbard Brook Experimental Forest site map with weather stations. Wind data recorded at Site 1, SCAN site, Kineo tower, MWS, and Pierce Lab (HQ). HBEF base map courtesy of Mary Martin.

The data we recorded revealed interesting characteristics of the cold air pool that formed that night. The temperature profile in Figure 2 reveals several different processes that impacted the evolution of the cold pool. A rapid cooling throughout the depth of the profile and especially near the ground (lowest lines) in the late afternoon signaled the formation of the cold pool (2100-2300 UTC). The separation of the lines indicates a cold air pool is present and that the air warms going up through the cold air pool – a typical characteristic. The data suggest that the top of the cold air pool was at about 145 meters above the ground – near the top of the tethered balloon – and remained there through the first half of the night.

Figure 2: Time series of temperatures at all tethered data loggers. Temperatures are filtered using a 10-minute moving average.

During the night, warmer air began to move in above the ridge tops. This warm air was first observed at the Kineo Tower along the south ridge; winds increased in association with the warm air between 0600-0800 UTC (Fig. 3). At about 0800 UTC (4:00 am EDT), this warm air mixed down past the height of the balloon, to the 135-meter temperature sensor (evidenced by the rapid increase in temperature of the top several lines just after 0800 UTC). This mixing down of warm air continued through 1000 UTC to about the 85-meter sensor – an erosion of 60 meters (~200 feet) of the cold air pool. (The cold air pool is gone where the upper temperature lines are nearly on top of each other.) Kineo Tower wind speed and temperature both decreased after ~0800 UTC revealing the pulse of warm air advection subsided. The undulating temperatures through 1200 UTC suggest a possible seiche (i.e., sloshing of cold air in the valley, like water in a bathtub); the decreasing wind speeds above the cold air pool (e.g., Fig. 3) supports this possibility.

Figure 3: Mount Kineo 15-minute averaged wind speed (solid black line), maximum wind speed (dashed line), and temperature (red line). Triangles along the horizontal axis represent time of sunset (filled; 2135 UTC) and sunrise (outlined; 1127 UTC).

All the data we collected suggest that the top of the cold air pool was maintained at ~145 meters (~480 feet) deep before the warm air began to erode it. We think that this is a typical maximum depth because of the shape of the orography. Just north of the tethered balloon the land rises, then plateaus at about 145 meters (~480 feet) above the valley (Fig. 4). In order for the cold air pool to deepen further, it would need to fill a much bigger volume above the plateau. In addition, while katabatic winds are supplying the valley with cold air, the cold air is also draining to the east into the Pemigewasset Valley, which balances the katabatic winds. If the cold air pool were to deepen, it would drain faster across a wider channel, especially over the low hills on the south side. This would reduce the cold air pool depth and drainage rate until the katabatic flow and drainage are back in balance.

Figure 4: Schematic diagram of drainage flows during maximum observed cold air pool height in HBEF. Small blue arrows (lower panel) represent katabatic wind flow. Large blue arrows (upper panel) represent mountain wind (drainage flow) of the cold air pool; the large orange arrow represents the increased drainage that would occur to the southeast if cold air pool height were to increase above the ridge height. Vertical black line shows location of vertical cross-section. Vertical transect of elevation generated from Google Earth (2017). HBEF base map courtesy of Mary Martin.

Understanding these dynamics helps advance understanding of where inversions typically occur on mountain slopes, energy and water cycling in the forest ecosystem, and how forest health may be impacted by these atmospheric conditions. Meteorologically, this research helps us improve forecasts of freezing rain, dense fog, and other hazardous conditions related to cold air pools.

Future Work

More observations are needed to definitely measure cold air pool height along the full west-to-east length of the valley. These will include using an array of temperature data from around the entire valley and more vertical profiles. Observations of cold air flow at more locations around the valley will help quantify katabatic inflow and mountain drainage outflow rates. Numerical modeling will be used to compare with observations and develop new hypotheses of how cold air moves within the valley and how disturbances, such as warm air advection, impact the stability and depth of cold air pools.

The ability of forecasters to have “real-time” data of moisture at the upper-levels, instead of the current system of twice daily soundings via balloon, will dramatically increase forecasting accuracy.