Barthram Night Cooling Toolv0.1

Hourly fall of temperature during the night · Barthram's method with the night cooling nomograms, Met Office Forecasters' Reference Book (1997) §2.4.2.3, Fig. 2.7

Date

Afternoon

Type of night (C)

Use a representative ascent: is there an inversion with its base below 900 hPa at the time of maximum? Will the night be best described as cloudless (radiation night) or 8/8?

Overnight

TR on the cooling curveTR is not used on cloudy or windy nights

In Auto mode, TR is not used when you choose an 8/8-cloud night, low or medium cloud is 6 oktas or more, or the gradient wind is 20 kn or more.

Options

Forecast

    Cooling curve

    Nomogram working

    Redrawn from Fig. 2.7, with the lines digitised from the scanned chart. Follow the orange construction: Tmax and Td → diagonal to the TR line → across to the gradient-wind line → down to the clear-sky minimum line → back to sunrise, then up by the cloud and summer corrections. The lower panels give the times and corrections.

    Step-by-step

    StepWorkingValue
    Method notes and limitations: read before relying on the output
    1. Method. Saunders' discontinuity temperature TR = ½(Tmax + Td) − C, where C depends on the type of night. The nomogram then gives a clear-sky minimum from TR and the gradient wind. Corrections for low and medium cloud (and, in summer, for the short night) are added to that. The cooling curve is drawn through Tmax in mid-afternoon (a monthly time, from 1330 UTC in December to 1545 UTC in June and July; local solar time is used as UTC), TR at the time of discontinuity, and Tmin at sunrise.
    2. Digitised chart. The wind lines, cloud-correction lines, the time-of-discontinuity and sunrise curves, and the summer correction were digitised from a scan of Fig. 2.7. Expect reading errors of about ±0.2 °C and ±5 minutes, which is similar to reading the paper chart by hand. Times are interpolated between mid-month values.
    3. Summer C for "8/8 cloud with an inversion". The summer chart's label says C = 1.0, but the dew-point scale printed on that row is offset by the amount you would expect for C ≈ 2.5. That would also fit the pattern of the other rows, where each summer value is the winter value plus 0.5. The tool uses the label (1.0) by default, and a tick box lets you switch to 2.5. The two choices give TR values 1.5 °C apart.
    4. Ranges. The winter wind lines are drawn for TR from about −10 to +14 °C. The summer lines are drawn for about +3.5 to +22.5 °C, and they curve near their lower ends. Outside these ranges the tool extends the lines in a straight line and shows a warning.
    5. Cloud corrections. The chart has separate graphs for medium cloud and low cloud. The tool adds both corrections together. The ≥ 20 kn line is used for the 20–29 kn and ≥ 30 kn wind classes. The summer short-night correction is only given for winds up to 19 kn, so it is zero for 20 kn or more.
    6. When to use TR. The text says to ignore TR on cloudy, windy nights but gives no thresholds. The thresholds used in Auto mode (8/8 night type, 6 oktas or more of low or medium cloud, or 20 kn or more) are this tool's assumption. You can override them. When TR is not used, the curve is drawn smoothly from Tmax to Tmin, and the shape between the fixed points is illustrative only.
    7. Times and location. The sunrise and discontinuity curves are in UTC and were drawn for the United Kingdom (about 52° N), with summer as April to September. The times do not carry over to other latitudes or to the southern hemisphere.
    8. Fog. The method applies to nights without fog. Once the temperature reaches the fog point, the real cooling usually slows sharply, so the curve after that time is unlikely to be valid.
    9. Snow cover. The book (§2.4.3) says that under clear skies and light winds the minimum over snow is likely to be 2–4 °C lower and to be reached earlier. With cloud it is about 1 °C lower, or less in winds of 15 kn or so. This tool simply takes 4 °C off and moves the time of the minimum to 0100 UTC, the time marked on the chart for a non-thawing snow surface. Treat all outputs as unreliable when there is snow cover.
    10. Not an operational product. Use alongside NWP, observations and forecaster judgement.