The health demonstrator work package led by Prof. Dr. Christian Witt aims to quantify the climate risk for vulnerable population groups and the health insurance sector in two different urban contexts (Central Berlin and Potsdam), representing two typical European cities. To this end, we link health damage functions (number of daily hospitalisations) to observed and simulated climate data. The focus in our showcase is on (urban) heat stress and air pollution affecting people with chronic respiratory diseases. Next to analysing and predicting climate risks, we demonstrate how specific counteracting measures can help to reduce the impacts for vulnerable population groups and thereby to alleviate the costs.
Building on experiences gained in previous projects, we quantify the sensitivity of a vulnerable population (e.g. chronic sick patients with a respiratory disease) to climate variability (heat stress, often amplified by air contamination). The product is a ‘health damage function’, linking environmental and air pollution data to impacts on the population (morbidity, represented by hospitalization) and healthcare costs. This health damage function will allow the public and private health insurers to assess the impacts (e.g. hospitalization costs) depending upon a climate event.
For the acquisition of health data, the team at Charité – Universitätsmedizin Berlin started cooperation with the hospitals Klinikum Ernst von Bergmann gGmbH in Potsdam, the DRK Kliniken Berlin, and the Vivantes Netzwerk für Gesundheit GmbH in Berlin. Data was collected from patients with chronic obstructive pulmonary disease (COPD) who were admitted to one of the participating hospitals between 2005 and 2017. In total, the assembled health database contains 54,396 anonymized patient datasets from Berlin and 2,222 datasets from Potsdam.
Scientists from ARIA Technologies, the Potsdam Institute for Climate Impact Research, and the Technical University of Denmark prepared the environmental database, which consists of air pollution and meteorological data. For the database, ground station measurements of the Deutscher Wetterdienst (German Meteorological Office) were spatialised on a regular grid with a resolution of 500 meters, re-gridded to postal code and aggregated by district. ARIA Technologies performed the spatialisation of pollutant fields by applying a land use regression-based approach. The temperature data was calculated by climate researchers from the Technical University of Denmark with a remote sensing-based methodology able to capture temperature spatial patterns within postal codes, including urban heat island effects. For this purpose, Landsat 8 satellite imageries were retrieved for the red, near infrared, and thermal bands with a resolution of at least 100 meters.
After consolidation of the health and environmental databases to a time series with daily values, we calculated the health damage functions for Central Berlin and Potsdam. The derived risk ratios are now being included into an Oasis-LMF prototype model. This prototype can then be tested by insurance companies. Also, datasets with future projections are being worked on, to predict future health damage under different scenarios of climate change.