Adapted from Preprints, 18th AMS Conference on Severe Local Storms, San Francisco CA, 19-23 February 1996.


Mark A. Darrow

and Michael L. Rehbein

National Centers for Environmental Prediction

Storm Prediction Center

Norman, OK


The National Weather Service (NWS) is currently in the process of modernizing its entire forecast program. As a result, the National Meteorological Center (NMC) has restructured its operations into several "prediction centers" across the nation. The National Severe Storms Forecast Center (NSSFC), once comprised of the Severe Local Storms Unit (SELS) and the National Aviation Weather Advisory Unit (NAWAU), have become part of the Storm Prediction Center (SPC) and Aviation Weather Center (AWC), respectively.

Part of the mission of the SPC is the forecasting of convection (both severe and non-severe) across the 48 contiguous United States. In recent months, the Day 1 convective outlook (AC) at SELS has been expanded. The current format of the AC includes separate discussions for both severe and non-severe convection at least four times daily.

Studies concerning deep convection in the West Coast region have generally been confined to cases involving severe weather development (e.g., Hales 1993). In an attempt to improve thunderstorm forecasting in the region, this study examines the meteorological conditions associated with the development of thunderstorms along the immediate Pacific coast (within 100 km of the coast). Upper air soundings from individual cases were used to construct composite soundings representative of the thermodynamic conditions associated with convective development. Data were collected from Quillayute, WA (UIL), Salem, OR (SLE), Medford, OR (MFR), Oakland, CA (OAK), and San Diego, CA (Miramar, NKX). The composite soundings were evaluated for the following: 1) convective available potential energy (CAPE), 2) lifted indices (LI), 3) environmental lapse rates, 4) moisture profile, and 5) wind profile.

In the past SELS forecasters have used an empirical rule to help predict thunderstorms along the West Coast during the cool season. This rule states that when the 500 mb temperature is equal to or less than -25 degrees C and significant synoptic scale upward vertical motion is forecast, thunderstorm development is possible. As part of this study, we will examine the validity of this rule.


Soundings were collected from days in which general thunderstorms (more than one thunderstorm reported per 3000 square miles) occurred along the immediate West Coast of the United States. This investigation spanned a two-year period beginning January 1, 1993 and included the "cool season" months of October through March. On a day in which thunderstorms were observed, the most representative sounding in space and time was chosen to represent the thermodynamic profile associated with this convection.

The meteorological parameters obtained from the individual soundings included height of mandatory pressure levels, temperature, dew point, and wind speed and direction. The mandatory level data were averaged to obtain a composite sounding for each location. Although the composite sounding technique will smooth some details in the profiles, the general sounding characteristics such as moisture depth and lapse rates will be retained. In order to obtain the composite wind field the total wind was broken into its u and v components. After averaging each component, an average total wind was obtained. Soundings were analyzed using SHARP 95 (Hart 1995).

In this version of SHARP, parameters involving temperature are computed using the virtual temperature correction. The reader is referred to Doswell and Rasmussen (1994) for further details of using virtual temperature in CAPE calculations.

Upon completion of each composite sounding, several thermodynamic parameters were derived. These parameters included: 1) CAPE, 2) LI, 3) 850 to 500 mb lapse rates, 4) 700 to 500 mb lapse rates, 5) K-Index, 6) mean lower layer (sfc to 800 mb) relative humidity, 7) mean mid/upper layer (800 to 250 mb) relative humidity.

Closely related to CAPE is the LI. In this study, we lifted a parcel from the boundary layer to various heights. Lifting parcels from a lower level to a higher level yields temperature differences between the parcel and its environment. If the parcel temperature is warmer than its environment, it is said to be unstable and its LI is negative.

Another index used to determine the stability of a given environment includes the K-index. This index uses temperatures at the 850, 700, and 500 mb levels as well as the dew points at 850 and 700 mb. According to the NWS Training Center (1989), a K-index between 15 C and 25 C supports isolated to widely scattered thunderstorm development west of the Rocky Mountains. However, operational experience suggests a K-index of 15 C to 20 C is rather low for thunderstorm occurrence, even in the western U.S.


The composite soundings for UIL, SLE, MFR, OAK, and NKX are shown in Figures 1-5, with data sample size and thermodynamic parameters derived from the composite soundings listed in Table 1.

Figure 1. Composite sounding for Quillayute, WA (UIL).

Figure 2. Composite sounding for Salem, OR (SLE).

Figure 3. Composite sounding for Medford, OR (MFR).

Figure 4. Composite sounding for Oakland, CA (OAK).

Figure 5. Composite sounding for San Diego, CA (NKX).

A striking feature of the soundings is the negligible CAPE (20 Jkg-1 or less). The minimal CAPE is consistent with the positive values of 500 mb LI's. Note that the 700 mb LI's are around zero, suggesting that the small unstable layer is below 500 mb. This indicates that the 700 mb LI is more representative of convective potential than the "standard" 500 mb LI.

Table 1. Shown in columns 2-4 are the total number of soundings for each site (Total #) and the breakdown by time (Total 12Z, Total 00Z). Columns 5-9 show the K-Index, temperature at 500 mb (T 500 mb), boundary layer to 700 mb LI (LI 700 mb) , boundary layer to 500 mb LI (LI 500 mb) ,and vertical totals (VT).

As seen in Table 1, 500 mb temperatures warm from north to south, consistent with the normal latitudinal variation of temperature. However, the two southern locations have 500 mb temperatures greater than -25 C during a typical thunder situation which is inconsistent with SELS' empirical rule. Despite the latitudinal variation in 500 mb temperature, the 850-500 mb lapse rates are similar at all sounding locations. The average 850-500 mb lapse rate (vertical totals) is 6.6 deg C / km (T at or above 27 deg C), while the 700-500 mb lapse rate is even steeper at 6.9 deg C / km.

Another common feature to all soundings is the presence of deep low-level moisture. Although actual mixing ratio values are less than 7gkg-1, average sfc-800 mb relatively humidities are quite high (greater than 80 percent). In contrast, relative humidities from 800-250 mb average around 35 percent. Since this is a potentially unstable environment it can be favorable for convective development. If we assume that d /df<0, i.e., that the layer is more humid at the base than at the top, the base will become saturated before the top in a rising motion. From that point it will follow the saturated adiabat while the top continues along a dry adiabat, which makes the layer absolutely unstable (Iribarne and Godson, 1986).

The composite thermodynamic profile described above yields "relatively" low K-indices. This is to be expected since 700 mb moisture contributes significantly to the K-index value. Such values might lead one to conclude organized convective development is not likely.

In general, winds are observed to have a westerly component throughout the entire sounding, which is a very common post frontal wind profile. This suggests upward vertical motion needed for convection is therefore not produced by low level warm air advection processes. However, when westerly winds traveling across the Pacific Ocean interact with land, frictional convergence/upward vertical motion increases. A west wind is also forced up the higher terrain, increasing the upward vertical motion of an air parcel.

With the nearly unidirectional wind profile, cold air advection is often limited to mid and upper levels. This allows for diurnal heating in the boundary layer which further steepens the environmental lapse rate during the day. As shown in Table 1, most soundings were taken at 00Z, suggesting a higher frequency of thunderstorms during the afternoon and/or evening hours. However, nocturnal thunderstorm activity is not uncommon along the immediate Pacific Coast. As mentioned above, trajectories across the eastern Pacific Ocean transport the characteristics of its airmass into the region and destabilization often occurs through warming and moistening of the boundary layer by heat and moisture fluxes from the ocean (Reed and Albright, 1986).


In this study, which spanned a two year period, soundings were collected from five Pacific coast upper air stations. An average sounding was created for each location from days in which general thunderstorms were reported within the vicinity of the upper air site. The composite soundings show the following characteristics: 1) minimal instability (700 mb lifted indices near zero), 2) steep lapse rates (vertical totals at or above 27 deg C), 3) potential instability, and 4) post frontal upslope flow.

In conclusion, some indices used to predict thunder may or may not be useful when dealing with convection along the immediate West Coast during the cool season. The best predictor for convection appears to be a post frontal environment where the thermodynamic profile is potentially unstable and upslope flow aids upward vertical motion.

Although the empirical rule often used at SELS is rather general, its inferences to steep lapse rates (-25 deg C at 500 mb) and upward vertical motion (upslope flow, differential cyclonic vorticity advection) do apply. However, the actual lapse rate (VT at or above 27 C, 6.6 deg C / km) and thermodynamic structure in relation to frontal systems seem to prove more useful than strictly using the 500 mb temperature and synoptic upward vertical motion combination.


The authors would like to thank Robert H. Johns, John E. Hales, Steven J. Weiss, and Stephen F. Corfidi for reviewing this manuscript and offering helpful suggestions to improve this paper. We also extend thanks to John A. Hart for graciously allowing us to use the newest version of SHARP before its official release. We are particularly thankful to the computer operators at the National Severe Storms Forecast Center for retrieving all the sounding data necessary for this study. Without their help, this study could not have been performed.


Doswell, Charles A., E.N. Rasmussen, 1994: The Effect of Neglecting the Virtual Temperature Correction on CAPE Calculations. Weather and Forecasting, 9, 625-629.

Hales, John E., 1993: Topographically Induced Helicity Enhancement and its Role in the Los Angeles Basin Tornado Maximum. Preprints, 17th Conf. On Severe Local Storms, St. Louis, Missouri, Amer. Meteor. Soc., 98-101.

Hart, John, 1995: SHARP 95. NOAA/NWS NSSFC Kansas City, MO.

Iribarne, J.V. and W.L. Godson, 1986: Atmospheric Thermodynamics, 259 pp.

NWS Training Center, 1989: The Skew-T Log-P Diagram, 85 pp.

Reed, Richard J., Albright, Mark D., 1986: A Case Study of Explosive Cyclogenesis in the Eastern Pacific. Monthly Weather Review, 114, 2297-2319.