Forecast Distribution

• The distribution plot shows the range of possible outcomes for continuous forecast variables, or probability density function (PDF). A continuous variable can have infinite different values within a range of values, e.g. precipitation  occurring in a day, e.g. 2.501, 2.587, etc . A probability density function (PDF) is an expression that describes the probability of a continuous random variable falling within a certain range.
• Depending on the models selected, you may see additional distribution curves, allowing you to compare Salient’s forecast to other models from NOAA/ECMWF and climatological forecasts. If you are viewing an anomaly, there are multiple reference periods available for calculating the anomaly (Salient climatology, 30-, 10, and 5-year norm). The selected reference period is appended to the x-axis label.
• The distribution curves show the sharpness of the forecast, indicating higher confidence about the range. You can visually see whether the Salient forecast and other government and climatology forecasts converge or diverge from one another.

Figure 1: The gray curve shows Salient’s proprietary 40 year climatology forecast for temperature. The Salient forecast (blue) predicts that temperatures will range from -1 C to a median above normal temperature within a narrow range compared to ECMWF SEAS5 (purple) and . The arrowhead symbols along the x-axis shows the medians of each forecast and the observed value from ERA5 Reanalysis (if you are viewing a past forecast).

• The y-axis of a probability density function (PDF) curve represents the probability density, not the actual probability. The probability of a variable falling within a certain range is obtained by finding the area of the curve on the left (below value) or the right (above value) of it.

Figure 2: The Risk View, using relative risk, highlights areas where the forecast probability exceeds the climatology probability. The probability of exceeding a value is obtained by finding the area under the curve to the left or right of the value.

Above-Below-Near Normal %

The Above-Below-Near Normal %” chart shows the probability of being above, below, or near normal relative to Salient’s climatology across multiple models. This chart shows whether models converge or diverge in their predictions of being in one of the three terciles. This alternative view of Forecast Distribution is available from the Forecast tab when the field is “Above-Below-Near Normal %”.

Figure 2: Above

v2/forecast distribution

https://api.salientpredictions.com/v2/documentation/api/#/Forecasts/forecast_distribution