Precipitation related × Precipitation parameters are under carefull testing at the time when you are reading this text. So far the results are good, but the methodology that works best for each region differs greatly based on the distance between the radar and each weather station and what radar we actually use(in a few cases, some farther radars work better than the ones close by). In order to quantify this possible spatial variations, we implemented 2 methods: one that tends to be more conservative and dry(labeled as DRY) and one that tends to be more water friendly and wet(labeled as WET). Use your judgement when using this until we find an unique solution that is reliable all the time. If it seems to you that it rained a lot outside(especially if you had a thunderstorm in your area), use the wet estimates. if not, use the dry ones. As we can't adjust the precipitation using rain gauges because ANM doesn't provide precipitation data freely, we need to improvise. We are sorry for the inconvenience. Anyway, a lot of countries are using multiple estimations to make out the full picture so consider this as a mini "ensemble" system. Individual-radar precipitation types use CSU S-band polarimetric hydrometeor classification. Where an ICON-EU thermodynamic profile is available, the existing wet-bulb melting-energy correction estimates how ice crystals, aggregates and wet snow change while falling toward the surface. To limit classification to pixels with meaningful precipitation, the layer is displayed only where the estimated instantaneous precipitation rate exceeds 0.1 mm/h. It describes estimated hydrometeor type, not a direct surface observation. As for everything, there are several limitations: At big distances, precipitation types become unreliable, as the radar beam goes too high up and it will almost all the time identify only snow or a mix. if you see sudden or nonsense transitions from rain to snow, that is a normal error. Precipitation types are highly dependent on the model thermal profile. if it is too warm/cold, the results might be wrong. This also cant be helped too much
