Parameters deduced from radiosonde data and numerical model data traditionally play a role in the forecasting of convective storms. Examples of such parameters are the lifted index, the K-index and the Boyden index. These parameters are generally defined in terms of temperature and moisture variables at different altitudes in the troposphere and can be calculated using either observational data or forecast data from numerical atmospheric models.
The skill of the various forecast parameters as predictors of thunderstorms varies greatly (e.g. Haklander and van Delden, 2003; Manzato, 2003). Other forecast parameters that have been developed do not address the likelihood of thunderstorms, but merely the overall threat of severe weather associated with convective storms. Examples are the SWEAT index and the index commonly referred to as the Craven Significant Severe index. There are also parameters that address the threat of large hail, severe winds or tornadoes specifically. For example, the Energy-Helicity Index (EHI) and the Significant Tornado Parameter (STP) were developed to be predictors of tornadoes.
Some weaknesses of some of the forecast parameters are
they may only work well in a particular area
they may keep a forecaster from really understanding what is going on
Examples of parameters that work only in a particular area are the K-index and the Total-Totals index, that incorporate the temperature and dew point temperature at 850 hPa in their respective formulas. That pressure level was probably chosen as one that is representative of the boundary layer. Across the central U.S. 850 hPa probably is, the surface generally being near 950 hPa in most places. At sea level, however, the 850 hPa level may be well above the boundary layer and therefore not very representative of the air that may flow into a storm's updraft.
Of course, the second point needs further clarification. The problem with many parameters is that they either do not represent anything physical or try to combine to many physical quantities with one formula. A particularly complicated example is the Severe weather threat index (SWEAT):
where f is the angle between the wind direction at 850 hPa and 500 hPa and TT is the TotalTotals index.
The index tries to incorporate instability by incorporating the TotalTotals index, which is another index that is based the temperature and dew points at various pressure levels. Additionally, the SWEAT-index takes wind shear into account by the last three terms. The effect of veering of the wind, as expressed in the last term is probably intended to take storm-relative helicity into account. This type of parameter gives the forecaster one value at each location, but does not give any information on instability or wind shear that a forecaster probably wants to inspect separately. Additionally, it still does not take the 'ingredient' of lift into account. If a forecast based mainly on this parameter turns out to be incorrect it will not be clear to say why it went wrong. It doesn't allow the forecaster to learn from past experiences.
It is therefore recommended that parameters are used only within the conceptual framework of an ingredients-based forecast methodology. Parameters can be used therein to assess the presence and magnitude of the several ingredients, e.g. there are parameters that assess lift and parameters that assess instability. A list of suggested forecast parameters is given in the next section.