Overview of Lapse Rate Research

By Karl Philippoff

As a weather observer and research specialist on top of Mount Washington, in addition to my usual observer duties such as taking hourly observations, releasing forecasts, thoroughly checking our daily observations and keeping our weather instruments calibrated and in good working condition, I also look into the data that we collect both on the summit and in other weather stations we have set up around the White Mountains to investigate how the data that we collect is connected to other research projects conducted in other mountainous areas. One of the projects that I have become involved in my time on the summit involves the calculation of near-surface lapse rates for Mount Washington over daily, monthly, seasonal, and annual time scales.

Before proceeding further, what, exactly, is a lapse rate? Any lapse rate in atmospheric sciences involves the change in a desired parameter (usually, but not always, temperature with a change in height). Diagnosing in the atmosphere is key parameter in determining the stability characteristic of the atmosphere, with greater changes in temperature with height corresponding to a greater degree of instability, and lesser changes with greater stability. Especially in the summer season in New England, the lapse rate is one of a few key parameters that meteorologists look to evaluate whether thunderstorms will form on a given day (along with low-level moisture and a lifting mechanism). These lapse rates are known as free-air lapse rates, as they are usually taken to be vertical profiles of temperature above a flat ground surface.

Measuring lapse rates close to surfaces, rather than in free-air, is important as this is where organisms, snow, and glaciers tend to be found. Determining the rain/snow line as well as accurate snow hydrological and ecological modeling depend on accurate temperature estimates along surfaces to determine an organism’s ecological niche, as well how much snow would be expected to melt at each elevation during the spring melt season. As surfaces can be variable (e.g. different vegetation coverage, face different directions (north-facing vs. south-facing, windward vs. leeward), it is not always trivial to extrapolate the vertical temperature profile measured in the free-atmosphere to that measured along a significant gradient, as from the base to the summit of Mount Washington. Temperature itself is also a key factor controlling many environmental processes, making estimations of this quantity useful for a variety of fields, not only for meteorology or climatology.

Usually this research is done by using stations in valley locations and then extrapolating the temperatures measured in that location to higher elevations by using what is usually called the environmental lapse rate. This lapse rate has been found to be the typical temperature change with height in the lower atmosphere across the globe. Right away, you should notice a few problems with this methodology. First, this lapse rate is the average rate of change in the free atmosphere, not along a sloped surface, so this ignores any consideration for the changes to this lapse rate that might occur due to varied surface characteristics. And second, that this is a globally averaged value which eliminates any contributions that may be caused by seasonality or different climates. As you might imagine, the lapse rate over a tropical rainforest and over the tundra in the winter may be quite different! While it may be acceptable as a first approximation, using data collected in the region of interest is always preferable, especially if it is collected over a substantial length of time to more accurately give a representation of its average and its variability.

In this particular study, I was looking at temperature changes along the surface between the base of the mountain and the summit using data taken at weather stations arranged at regular intervals, roughly every 1,000 feet, along the Mount Washington Auto Road. These stations are all measuring temperature at the same height above the local ground surface. If you’ve ever seen the Auto Road Vertical Temperature Profile displayed on the left side of our current conditions page, it is a real-time glance at how these temperatures are changing with elevation every minute of every day (Figure 1). Additional benefits of using this data include:

· It consists of 7 full years of data (Jan 1st 2016- Dec. 31st 2022) while most other comparable studies use only 1 to 2 years of data.

· The stations are consistently facing toward the east-facing side of the mountain within short distance of each other, which minimizes discrepancies based on different aspects, exposure to winds (prevailing from the W to NW) and that experience similar weather at about the same time with regard to frontal systems.

· Data availability within the study period is excellent, with most stations having >99% data availability every minute over the full period.

· No other data set over a comparable length time at this kind of time resolution exists for the Northeast, making this dataset unique.

Figure 1: Current Conditions page highlighting the Auto Road vertical temperature profile section which displays a real-time look at the temperature measured by each weather station along the Auto Road, the same studies that were used in this lapse rate study.

The lapse rates reported in the literature usually determined with respect to daily maximum, minimum, and average temperatures, so each of each station’s maximum, minimum, and average temperatures were calculated for each of the auto road weather stations. Then a best-fit line was fit through the remaining stations to determine the daily lapse rate with respect the three different temperature data points. Once the daily lapse rates were calculated, the data was averaged to determine daily, monthly, seasonal, and overall averages.

Some of the more important results we found were:

The daily data, as grouped by season, was organized into a histogram, with every day being shown. Each showed that there was substantial variation around the average lapse rates for each season, with the greatest amount of variation seen in the winter season, and narrower, more peaked distribution during the summer season. Both the spring and fall seasons showed distributions somewhat in between either of these two extremes (Figure 2).

Figure 2:Histogram showing the distribution of lapse rates measured in each of the four meteorological seasons. Winter is defined as December, January, February (DJF), Spring as March, April, May (MAM), Summer as June, July, August (JJA), and Fall as September, October, November (SON). The y-axis on each of the subsetted figures is the number of days that calculated lapse rates fell within a given range of values (shown on the x-axis). 

All seasonal lapse rates were found to be significantly different from the environmental lapse rate, making it a poor estimator of temperature with increasing elevation in the White Mountains (Figure 3).

Figure 3: Vertical bar chart showing the variation in the lapse rate by season. Notice how the maximum temperature lapse rates always exceed the minimum temperature lapse rates, and that the minimum temperature lapse rates do not vary substantially by season, other than displaying a slight peak in spring. The dotted line refers the environmental lapse rate of 6.5°C/km, which is a global annual average, irrespective of local climate or season. Note that this value does a relatively poor job of estimating the calculated lapse rates over Mount Washington, especially in winter. 

The overall average lapse rate of 5.5 °C/km was in good agreement with prior studies of lapse rates in the White Mountains and greater New England (Figure 4).

Figure 4:Overall temperature and lapse rate averages including all the weather stations along the Auto Road and the summit station. The slopes of each of the lines represent the average lapse rate measured for minimum (blue), average (yellow-green), and maximum (purple) temperatures. The R2 values refer how well the best-fit line matches the station data, with values closer to 1 indicating a better fit.

When the daily data was grouped by month, this began to show somewhat of a dichotomy between cold season and warm season patterns. In this case the ‘warm season’ begins in May and lasts until September and is characterized by steep lapse rates, especially of the maximum temperatures. Between November and March, the lapse rates of the maximum, minimum, and average temperatures are much closer together and they also vary in concert, though maximum temperature lapse rate always exceeds the minimum temperature lapse rate. March and October are transition months that exist somewhere in between these seasons, with a larger spread between the maximum and the minimum lapse rates than the ‘cold’ season, but a smaller spread than the ‘warm’ season. This flip between ‘warm’ and ‘cold’ modes becomes even more distinct when smoothing the daily lapse rates using a 10-day running mean (Figure 5). During the warm season, the minimum, maximum, and average lapse rates are spread well apart, with very little crossover between three different lapse rate flavors. During the cold season, however, the variation between the three flavors is much more tightly constrained, and they are also substantially more variable.

Figure 5: 10-day running mean of the calculated daily lapse rates using the Auto Road weather stations. This figure displays the running means of the maximum, minimum, and average temperatures measured at each station, and then using a best-fit line to find a singular lapse rate between the base and the summit. The red rectangle highlights the ‘warm’ season, while the blue rectangle highlights the ‘cold’ season.

While this particular phase of the project has been completed, there are still multiple avenues for future research concerning lapse rates within the White Mountains. These include:

· Categorizing lapse rates according to precipitation events, especially those close to freezing, to better determine the evolution of the rain/snow line during precipitation events for future higher summits forecasts and avalanche forecasts.

· Expand the weather station network to include stations on the westward or windward slope of Mount Washington to determine the role of wind exposure on the temperature profiles measured, and strengthen the interpretation of the existing data.

· Calculating the lapse rates with respect to the Pinkham Notch COOP weather station so that lapse can be estimated back into the 1930s.

To learn more about the team’s lapse rate research, visit mountwashington.org/research/current-research-projects/lapse-rate-research/

Seek the Peak, our largest annual fundraiser, is underway! Learn more at seekthepeak.org or consider making a donation to the MWOBS Weather Observers’ team page to support our work in forecasting, education, and research.

Find Older Posts