A few years ago(sometime in the late 90's) I came up with a simple technique for assessing
    instability on the front range without the assistance of upper air data. A lot of information about the
    upper air can be determined through the appropriate use of surface data. The best way to show
    how this technique works is through a case study example. I am choosing June 25, 1942 as my
    example, not out of  convenience, but since I just happen to be studying this case at the present time.

    On June 25, 1942, a thermal low was situated in the northern Texas panhandle, with a nearly
    stationary frontal boundary stretching from central Wyoming into central Colorado and then east into
    the panhandles and finally north and east into eastern Kansas and Missouri. North of the boundary,
    moist upslope flow was noted across northern Kansas, Nebraska and eastern Wyoming. Another
    area of surface low pressure(probably a thermal low) was located in the northern Rockies. A rather
    diffuse surface dryline was located across west Texas. A cold front stretched from eastern Idaho
    into western Utah. Ahead of the front, a very warm plume of air covered much of the central and
    southern Rockies.

    By midday(1230 MST), surface temperatures under the warm plume included 77F at
    Fort Bridger(7040 ft~782mb), 78F at Rock Springs(6760 ft~790mb), and 86F at Denver
    (elev. 5300ft~830mb). These temperatures lie nearly along the same dry adiabat on a skewt-logP
    diagram, meaning that surface potential temperature was fairly uniform from Denver to Rawlins to
    Rock Springs. With mostly sunny skies and brisk south to southwest surface winds, I assume that
    the thermal profile above these stations are close to dry adiabatic from the surface to 500mb. One
    can infer south to southwest upper level flow over this region from these surface winds. In such a
    flow regime, a deep elevated mixed layer would most certainly move downstream and cap the
    boundary layer over Douglas and Cheyenne, WY.

    Now simple take the midday surface temperature at Rawlins(the most immediate station
    to Douglas where surface data is available in the moist air) and move this up the dry adiabat to
    500mb on a skewt. This should be very close to the actual 500mb temperature. Why do I use
    midday surface data at Rawlins? Because in my experience, the thermal profile is closest to
    adiabatic from the surface to 500mb at this time. In the mid to late afternoon, a strong
    superadiabatic layer often develops near the ground. So using data later in the day would
    give a 500mb temperature that is too warm. Now take the surface temperature and dewpoint
    at Douglas, WY(elev. 4900ft or around 842mb) and lift this parcel up to 500mb(dry at first and
    then moist). Now subtract this parcel temperature from the estimated 500mb temperature above
    Rawlins and this yields a 500mb lifted index of around -10 or -11. There was in fact a killer
    tornado near Wheatand, WY that evening.

    This technique works well during the front range severe weather season(June-August) when
    500mb temperatures are slow to change, especially over a span of a few hours. Also, this is
    the time of year when the elevated mixed layer is a given when southwest or westerly flow
    is present across the central and northern Rockies. Of course, severe local storms are rare
    on the front range from October to April anyway.