The PhD defence will be held in Auditorium 3 in The ZEB-building, and the trial lecture in Aud 1 in the Geology Building. In some cases, it will be possible to attend the trial lecture and dissertation digitally, in that case a link to Zoom will be posted.
Trial lecture
Thursday 3 September, 10:15-11:00, Aud 1, The Geology Building:
Precipitation Parameterizations in Earth System Models
Conferral summary (in Norwegian)
Snø oppstår høyt oppe i atmosfæren før den faller ned som snøfall. I sin avhandling undersøker Franziska Hellmuth hvordan snøfall dannes og hvordan disse prosessene representeres i vær- og klimamodeller. Ved å koble avanserte bakkeobservasjoner med bilder av snøfnugg viser studien at tradisjonelle radarmetoder gir stor usikkerhet for snøfall nær kyststrøk og at værvarslingsmodeller overestimerer overflatesnøfall, og underestimerer snømengden i atmosfæren for en stasjon på fjellet på grunn av mikrofysiske feil i parametrifiseringer. Satellittmålinger viser at klimamodeller lider av et "konstant yr"-problem i «kald versjon» ved å oversimulere for ofte og for lett snøfall fra skyer med underkjølt vann på global skala, noe som påvirker modelleringen av jordens albedo og energibalanse.
Main research findings
Popular scientific article about Hellmuth’s dissertation:
Bridging the Scales of Snowfall: from Snowflakes to Global Climate Models
Snowfall is a vital freshwater resource and a major planetary cooling shield, yet representing it accurately in weather and climate models remains a major challenge. This thesis combines snowflake imaging, ground-based precipitation radars, and global satellite observations to evaluate atmospheric models across multiple scales.
In her doctoral study, Franziska Hellmuth shows that single-frequency radar retrievals are highly ambiguous because rapid microphysical transitions between dense graupel and aggregates change particle mass and fall speed independently of reflectivity, using snowflake images and vertical radar at a coastal Arctic site. Scaling up to regional weather models, we find that the they struggle to accurately represent clouds containing liquid and ice and the snowfall at the ground over complex terrain. Specifically, the models overpredict surface snowfall by up to 59% but underestimate atmospheric snow aloft by over 500%, showing that surface-only evaluation masks vertical biases. On a global scale using spaceborne sensors, we revealed that ten climate models suffer from a cold version of the “perpetual drizzle” problem, overestimating the frequency of supercooled liquid clouds and producing light snow far too continuously.
Because the cloud and snowfall biases directly distort the models ability to represent simulated solar radiation and surface warming, the study suggest to correcting them as this is essential to reduce uncertainty in future polar warming and climate projections.
Photo and other information:
Press photo: Franziska Hellmuth, portrait; 1500px. Photo: Private
Other photo material: Photo with description and credit as specified in the article above, size 2000px.
