Decoding the nomenclature for COMPASS experiments:

Each COMPASS simulation experiment occupies a unique spot in an eight-dimensional parameter space. Thus, eight numeric descriptors are needed in identifying each simulation. For mnemonic purposes, we also include an alphabetic identifier for each parameter. Thus each experiment name consists of a series of eight pairs of characters, with the first member of each pair being the alphabetic character suggesting the name of the parameter, and the second being the representation of the numerical value assumed by that parameter in that particular experiment. While this nomenclature design may at first seem cumbersome, we have concluded that it is the simplest way to assign unique names to each experiment. With just a little experience, we have found it quite easy to describe the various subregions of the large parameter space using our nomenclature. For example, discussion of "p3 versus p6" is our shorthand way of saying we are discussing how the simulations behave as the "p" parameter (which is the designation for environmental precipitable water, a good proxy for the temperature at cloud base) varies, all other parameters being held constant.

In most cases, we have opted to do experiments having only two values of each parameter, one large, the other small. The values are chosen to be realistic, but different enough so that any real convective sensitivity to the changes in the parameter will be evident in the simulations. In a few cases (such as "e" and "c") we have decided to add a third, "medium" value of the parameter, so that commonly encountered values of those parameters will be given proper consideration. In the case of the "m" and "n" parameters (which regulate the shapes of the buoyancy and vertical shear profiles, respectively), we have opted to use different paired values in the different parts of the "e" (CAPE, see below) space. This is because of some fundamental relationships between the way the "m" parameter can vary as "e" varies. To be more precise, when "e" is small, there exists a greater range of possible physically realizable values of "m" than at large "e"; one could use the same values of "m" for all values of "e", but they would have to be restricted to only those values that are physically realizable at large "e." It turns out that interesting sensitivities are evident for large values of "m" when "e" is small, and we have chosen to let our pairs of values of "m" increase at small "e" to allow for proper exploration of these important sensitivities. In physical terms, this means that when CAPE is small, it is possible to generate soundings with lower altitudes of maximum buoyancy, as compared to what happens when CAPE becomes very large, where flexibility in shaping the buoyancy profile starts to get lost. Note that although it is the relationship between the largest value feasible for "m" as "e" decreases that dictates our choices of "m" as a function of "e," we also have decided to let the shear shape parameter "n" assume the same values as "m" for each value of "e" studied. This is done partly for convenience, and partly because observational data suggest that larger values of "n" do indeed tend to be correlated with lower values of "e."

What follows is an example experiment name, and instructions on how to decode it. With the information below, it is possible to decode the structure of each experiment's design. In the graphical results section, we always add each experiment's full name to each image, so that the results are unambiguously attached to the proper experiment.

In the legend information below, the following abbreviations are used:

  • CAPE = convective available potential energy, a quantitative measure of the maximum bulk amount of vertical kinetic energy updraft parcels can acquire from their environment; updrafts seldom manage to convert all the CAPE into vertical kinetic energy, because of adverse effects of certain combinations of the other seven environmental parameters; one of COMPASS's main interests is in seeing which combinations of parameters lead to the most efficient convective overturning;

  • LCL = lifted condensation level, or the level of cloud base, dictated by surface temperature and moisture and boundary layer lapse rate;

  • LFC = level of free convection, or the level above which ascending parcels begin to experience positive buoyancy relative to the environment; this level can never be below the LCL due to physical realizability constraints.

Note that although the table below gives three possible values of LCL and LFC height, we are currently using only the smallest and largest values. Note also that although the experiment framework readily allows for either curved (semicircular) or straight hodographs, curved hodographs are used exclusively in the first phase of this work.

Sample experiment name = e2c2m2n4k2f6p6h9

CAPE parameter e:
1 = 800 J/kg
2 = 2000 J/kg
3 = 3200 J/kg

Hodograph radius parameter c (or s):
c = curved
s = straight
1 = 8 m/s
2 = 12 m/s
3 = 16 m/s

Buoyancy shape parameter m:
1 = 1.56 -> Zb' = 9.30 km
2 = 1.88 -> Zb' = 7.71 km
3 = 2.38 -> Zb' = 6.09 km
4 = 3.22 -> Zb' = 4.50 km
5 = 5.00 -> Zb' = 2.90 km
where Zb' is the altitude of maximum buoyancy, relative to the LFC

Wind profile shape parameter n:
1 = 1.56 -> Zv' = 9.30 km
2 = 1.88 -> Zv' = 7.71 km
3 = 2.38 -> Zv' = 6.09 km
4 = 3.22 -> Zv' = 4.50 km
5 = 5.00 -> Zv' = 2.90 km
where Zv' is the altitude of maximum v-wind, relative to the LFC, in an assumed curved hodograph situation; for straight hodographs, a curved hodograph is first constructed, then unfolded into a straight line, ensuring that the shear profiles are identical for both curved and straight hodographs with similar specifications.

LCL height (actually mixed layer depth) index k:
2 = 0.5 km
4 = 1.0 km
6 = 1.6 km

LFC height (actually moist layer depth) index f:
2 = 0.5 km
4 = 1.0 km
6 = 1.6 km

Precipitable water (PW; implemented as LCL Temperature) parameter p:
3 = 30 mm ( -> T_LCL = 15.5 C for k = 2)
6 = 60 mm ( -> T_LCL = 23.5 C for k = 2)

Midtropospheric relative humidity parameter h:
5 = 50% everywhere in the troposphere above the LFC
6 = 60% everywhere in the troposphere above the LFC
7 = 70% everywhere in the troposphere above the LFC
8 = 80% everywhere in the troposphere above the LFC
9 = 90% everywhere in the troposphere above the LFC

Using the above legend, the sample experiment name "e2c2m2n4k2f6p6h9" means that CAPE = 2000 J/kg, the radius of the curved hodograph = 12 m/s, the buoyancy shape parameter = 1.88, the shear shape parameter = 3.82, the LCL level = 2 (0.5 km), the LFC level = 6 (1.6 km), the precipitable water = 60 mm, and the relative humidity = 90%.