Regional climate over Northern Europe: from observations to high-resolution modeling

Abstract

Regional and global climate models are important tools to study the past climate and estimate the impacts of future climate change. Climate models can also provide input for other models that simulate, for instance, hydrological cycle or road weather. Recently, running climate models with fine grid spacings (< 4 km) has become affordable at climatic scales (10 years or more) due to increased computational resources. With such grid resolutions, deep convection can be resolved explicitly leading to an improved representation of heavy precipitation. Heavy precipitation events can cause major environmental and socioeconomic hazards due to flooding, landslides, and erosion, and therefore, their accurate representation in climate models is crucial. It has been shown that extreme precipitation events have become globally more frequent over recent decades as a result of global warming, and they are expected to intensify further in the future due to climate change. Motivated by the expected better representation of heavy precipitation compared to previous methods, high-resolution regional climate model simulations covering 1998–2018 were performed for the first time over the Nordic region with a regional climate model, HARMONIE-Climate (HCLIM). In this thesis, the skill of HCLIM in representing the features of the present-day climate was evaluated by comparing the model simulations to several observations. Moreover, the applicability of the HCLIM data to drive a road weather model, RoadSurf, was investigated. Because running high-resolution climate models is computationally expensive, the added value provided by such models needs to be quantified. Therefore, this thesis assessed the benefits of a high-resolution HCLIM setup with explicitly resolved deep convection at 3 km grid spacing over a setup with 12 km grid spacing and deep convection parameterization. In addition, past trends in observed extreme precipitation between 1901 and 2020 were investigated in order to put the future trends in context. The results of this thesis indicate that precipitation extremes have intensified in the Nordic-Baltic region. Extreme events also occur later in the year compared to the beginning of the last century. Precipitation extremes and other present-day climate characteristics over the Nordic region were well captured by HCLIM. In addition, the HCLIM-driven RoadSurf model demonstrated a good skill in representing road weather in the region. The high-resolution HCLIM setup was shown to improve especially high-intensity sub-daily precipitation events in line with studies conducted over other regions. The results support the use of HCLIM and RoadSurf models to produce climate change impact projections for the Nordic region. Furthermore, the results of this thesis emphasize the need for high-resolution convection-permitting regional climate models to reliably simulate high-intensity precipitation events.