WP1: Exposure

The objective of WP1 Exposure is to obtain and produce health-relevant exposure data for both historical and future time periods using a combination of open datasets, observations and several models. The parameters fall into three classes: a) climate data, b) other environmental data, and c) socio-economic data. The data are processed in collaboration with WP2 into indices and maps with potential health-relevance, and their actual connections to health are explored in more detail in WP2. This information feeds to WP3, where the indirect impacts of climate change on exposure and vulnerability through climate and urban policy are investigated, and finally to WP4 for GI synthesis. WP1 is divided into five tasks.

In Task 1.1a, we collect two observational datasets: 1) Continuous spatially resolved (10 km resolution) daily country-wide time series of a selection of variables with potentially high health relevance include e.g. near-surface air temperature, daily maximum near-surface air temperature, snow cover, precipitation, and relative humidity from an observation database managed by the FMI, and 2) even more detailed meteorological TURCLIM dataset from Turku, one of the Six Cities (see Section 2.2 for more details).

In Task 1.1b, we use regional climate models to provide an even more holistic picture of the past climate conditions and acquire variables that are not well presented in the observations (e.g., surface radiation and cloud cover). First, we use regional climate model simulations with the high-resolution (12 km) HCLIMALADIN model (Lindstedt et al., 2015). HCLIM simulations are partly based on the ongoing Nordic collaboration in the project NorCP. The historical period covers the years from 1980 until 2019, and Era-interim reanalysis data are used as the model boundary conditions to get an accurate description of the past climate. Furthermore, HCLIM-ALADIN will be used to get future climate projections under low (RCP2.6) and high emission (RCP8.5) scenarios for the period 2041-2060. To get an estimate on model uncertainty and on natural variability, the HCLIM simulations are complemented by acquiring and processing future climate model simulations from a model ensemble from EURO-CORDEX initiative. HCLIM-ALADIN and EURO-CORDEX results will be carefully evaluated and bias-corrected (Räisänen & Räty, 2013) to ensure their reliability for health-impact studies. The data from the regional climate model simulations will be used as an input for three complementary modelling approaches in Tasks 1.1c-e.

In Task 1.1c, HARMONIE-AROME weather prediction system (Termonia et al. 2018, Bengtsson et al. 2017) will be used for city-scale (resolution of 500 m) simulations to investigate relationships between climatic parameters and urban characteristics during heat waves. We will simulate at least one historical heat wave event for each of the Six Cities to get a detailed view on spatial heat stress expressed as universal thermal climate index. We will also study the impact of urban infrastructure on the heat waves by simulating an intense heat wave in Helsinki region on city-scale using one of the future scenarios with the current city-plan, plus with a future city-plan including adaptation to changing climate.

In Task 1.1d, we do road weather modelling with RoadSurf model (Kangas et al. 2015, Toivonen et al. 2019) to produce health-relevant data on road conditions for both past and future climate conditions.

In Task 1.1e, to complement physical simulations, we will use GIS data based spatio-statistical modelling methods, such as BRT, GLM, linear regression, and GA (Hjort et al. 2016; Suomi and Käyhkö 2012) to study spatio-temporal characteristics of UHI and relevant environmental factors, such as land use, topography and water bodies. The models will be performed on fine spatial resolution, in 100-500 m grid. Model calibration and validation will be based on the observations of TURCLIM (Turku Urban Climate Research Group at UTU GEO) local climate network in Turku. The model will be applied in the Six Cities to predict the intensity and spatial characteristics of the current and future UHI. The models are applied for long-time average conditions, for different seasons and for short-term extreme situations with increased health risks.

In Task 1.2 we track past and current patterns of physical living environments with LULC datasets (CORINE, SLICES) and remote sensing data (Landsat, Sentinel). Task 1.3 comprises of tracking and analysing the patterns and changes in social living environments using gridded socio-economic GIS data (YKR, 250 m grid; see section 2.2). Furthermore, we project future spatiotemporal changes in physical and social environment at high spatial resolution with the help of past and current environmental exposure data, city plans and participatory scenarios of potential urban development and statistical modelling (Rohat et al. 2019).