Previous: Method Next: Conclusion Up: Ext. Abst.

Discussion

Atmospheric waves have many generation mechanisms. Shear and convection are amongst many others like mountains, turbulence fronts or geostrophic adjustment. In this chapter techniques have been described to indicate the existence of both gravity and inertial gravity waves and different wave generating mechanisms in radiosonde ascents. The association of wave generating mechanisms with gravity waves from radiosonde ascents is a very difficult task due to noise in the data and the uncertainty in the mechanisms.

These techniques have been combined in a new algorithm which allows an automated analysis of wave events. A significant innovation is the use of the Stokes parameter method. The Stokes parameter spectrum has been developed for the use in single ascents and was used here to routinely find inertial gravity waves in radiosonde ascents.

35261 profiles were used for the analysis. Chi-square analysis introduced significance to the indicators and a method has used to describe associations between those indicators.

The results of the climatological analysis are illuminating. Most important of these results are

1. Convection (as identified using the CAPE criterion) is present
   • with tropospheric gravity waves (as identified using the ascent rate criterion) in 4.20% of 8898 summer profiles and in 5.88% of 9530 winter profiles.
   • with stratospheric gravity waves in 1.56% (summer) and in 9.08% (winter).
   • with tropospheric inertial gravity waves (as identified using the Stokes parameter criterion) in 6.05% (summer) and in 6.77% (winter).
   • with stratospheric inertial gravity waves in 6.19% (summer) and in 8.58% (winter).
2. The chi-square analysis shows that convection is significant only for tropospheric gravity waves and for stratospheric inertial waves in summer and gravity waves in both the troposphere and the stratosphere but not for inertial gravity waves in winter.
3. An upper limit to the % of gravity waves produced by convection can be estimated from the % of cases where significance indicators occur together to be 4.20% in summer and 9.08% in winter. These percentages should be compared to the total gravity wave cases: 16.14% in summer and 57.49% in winter. The proportion of gravity waves for which convection could be an important generating mechanism is therefore: 26.0% in summer and 15.8% in winter.
4. Convection and gravity waves appear together more frequently in winter than in summer, but convection and inertial gravity waves have no seasonal trend.
5. There is no latitudinal trend for gravity waves or inertial gravity waves together with convection, even though there is for convection alone.
6. Chi-square shows that shear is significant for all types of waves apart from inertial gravity waves in the stratosphere due to stratospheric shear. In summer it is true apart from stratospheric inertial gravity waves due to tropospheric shear and in winter it is true for tropospheric inertial gravity waves due to tropospheric shear and stratospheric inertial gravity waves due to stratospheric shear.
7. Maximum numbers of shear generated inertial gravity waves are 20% in the stratosphere and 18% in the troposphere. For short period gravity waves it is 16% in the stratosphere and 15% in the troposphere.
8. Both gravity waves and inertial gravity waves appear much more frequent in winter together with shear than in summer.

For the interpretation of the above results the following uncertainties must be borne in mind:

1. CAPE as an indicator for a convective wave-generating mechanism:
   • Relationship between CAPE and convection needs to be quantified through further work
   • Spatial distribution of CAPE: Strong convection is localised and radiosonde ascents may not be indicative of important regional convective activity
   • Temporal relationship with waves: Uncertainty in the wave generating mechanism means that a temporal delay between the occurrence of convection and that of waves may be important.

   These points may lead to an underestimate of convection.

2. Lack of indicator for orographic wave generation: It is possible that where convection is present, and shear is present, waves detected can nevertheless be of orographic origin. Orographic origin, however, can only be excluded by individual case studies and could therefore not be included in this study.
3. Choice of thresholds: The threshold values were chosen to be high enough that there would exist cases for each of the categories. It turns out that seasonal changes are well illustrated by this choice of threshold values.
4. The chi-square significance measure allows only a quantitative interpretation.

Despite these uncertainties, the study gives important new information about the incidence of gravity waves in the UK, and their relationship with local environmental factors. The indication from this study is that convection may play an important role in gravity wave production even in the mid-latitude winter.