Gálvez-Davison Index calculation

The GDI is a stability index that performs best on the tropics and subtropics. It works best in trade wind
regimes, and regions downwind such as eastern continents in tropical and subtropical latitudes. The GDI
considers only moisture and temperatures at four levels. It can thus be applied into different databases
such as analyses, model output, satellite-derived fields and soundings, for example.

Input data
The GDI requires temperature and mixing ratio data at four levels: 950, 850, 700 and 500 hPa. When
plotting the GDI in grids, an optional correction is available to improve visualization over high mountain
ranges, where high fictitious GDI values occur. If you want to apply this correction, you need surface
pressure data.

General philosophy
The index considers working with three layers as indicated in Figure 1. The data is then organized into
these three layers (for details just follow the equations presented later in this document). Layer A
considers 950 hPa data representing near-ground conditions. Layer B considers an average of 850 and
700hPa data representing the low-mid troposphere, and layer C considers 500 hPa data representing the
mid-level troposphere. You can find these layers in Figure 1, which also shows the conceptual model of
convective regimes in tropical/subtropical trade wind regions and their signature on the field of equivalent
potential temperature.

Figure 1. Conceptual model of dominant tropical convection regimes (white contours for clouds) and their signature in the
field (light gray contours, data in K). The three layers used for GDI calculations and corresponding pressure levels are also
included.
Calculation Algorithm
The first step is to compute potential temperatures [in K] and mixing ratios [in kg/kg] for each layer via
equations 1.1 through 1.6. Input temperatures are always in [K]:
(1.1)
(1.2)
(1.3)

(1.4)
(1.5)
(1.6)

Equivalent Potential Temperature Proxies (EPTP) are then calculated following one of the methods
described by Davies-Jones (2009). These are called “proxies” because they are not true equivalent
potential temperatures. A modification was made for simplification. It was noted that computing the
temperature at the lifting condensation level was potentially complex and using instead 850 hPa
temperatures produced negligible changes in final GDI values. Thus the formulas we used contain 850
hPa temperatures [K] as follows:

(1.7)

(1.8)

(1.9)

where [K] is an empirical adjustment constant, [J kg-1] is a latent heat constant

A dimensionless EPTP Core Index ( ) is then calculated. The is an enhancement factor that
addresses the potential for convection developing into the mid-troposphere, which is often sufficient for
convection to progress into the upper tropical troposphere as the mid-lower troposphere is where the most
stable layers often linger. A Mid-level EPTP factor ( ) and a Low-level EPTP factor ( ) are first
calculated via (1.10) and (1.11):

(1.10)
(1.11)

where [K] is an empirical constant that enhances EPTP variability. The is then calculated
via (1.12):

(1.12)
where [K-1] is an empirical scaling constant. The reasoning behind (1.12) is that if
significant amounts of heat and moisture are available at mid-levels, and large values are also present at
low-levels, the entire column is moist and the moisture is ground-based. This causes the potential for
convection to increases non-linearly based on the departures of and from the threshold.

A Mid-level Warming Index ( ) is then calculated. The accounts for the modulation of mid-
level stability by the presence of a warm ridge (increase in stability) or a cool trough (decrease in
stability). It is an inhibition index that only produces negative values as a function of how much warmer
are 500 hPa temperatures from the [K] threshold (~ -10 C). It can be calculated via (1.13):

(1.13)

where [K-1] is an empirical scaling constant and is the 500 hPa temperature in [K].

An Inversion Index ( ) is then calculated as inversions play an important role in inhibiting tropical
convection. The is an inhibition factor as well (i.e. only negative values are allowed) and is calculated
by considering two factors: stability across the inversion and dry air entrainment once convective cells
penetrate the inversion layer. The Stability factor is calculated via a simple difference of the
temperatures at 950hPa ( ) and 700 hPa ( ) in [K]. The 950 hPa level was used since the inversion
can sometimes be located at levels as low as 925 hPa, while other times rises up to 700-750 hPa. is
calculated via (1.14):

(1.14)

where [K-1] is an empirical scaling constant. A Drying factor is then calculated based on the
difference of EPTP between layers A and B via (1.15). In the tropics, dry air associated with subsidence
inversions has a strong signature in the EPTP field resulting in a sharp decrease of EPTP with height.

(1.15)

Both and are combined via (1.16) to calculate the :

(1.16)

The GDI can then be calculated by simply adding the three indices calculated:

(1.17)

An optional correction is available to improve visualization around high mountain ranges. NOAA
products online and Wingridds GDI equations use this correction. It is recommended when using GDI as
a gridded forecasting tool, since it will tamper the excessive fictitious values that occur over tall
mountains such as the Andes, Rockies, Mexican Sierras, Central American ranges or Hispaniola’s ranges.
This correction considers surface pressure in [hPa]. Note that the GDI is not truly applicable over
terrain located over 950 hPa due to the need of data at this level. Yet, validation has shown that GDI
variability in some of these regions still leads to high correlations, some of which even exceed r2=0.5. The
surface pressure correction is:

(1.18)

where [hPa], [hPa] and are all empirical constants tuned to improve GDI
visualization over elevated terrain.

A version of the GDI optimized for visualization can be attained via (1.19):

(1.19)