This dissertation empirically estimates the impact of climate variables onto economic outcomes by expanding existing models in three dimensions: Climate Measures, Space and Time. First, a range of extreme climate and drought indicators are introduced into a global panel dataset to evaluate their influence onto annual country-level GDP per capita growth beyond annual averages of temperature. Second, to assess the value of increasing the frequency of measurement, the analysis is conducted quarterly rather than annually. Third, to avoid country-level aggregation, nightlight imagery is used as a proxy for local economic activity at a higher spatial resolution, allowing local climate impacts to be estimated. Results suggest that the addition of extreme weather indicators adds significant information to the econometric analysis at both country and gridded level, although at the cost of increased complexity. These results can help to empirically calibrate Integrated Assessment Models and
provide more realistic climate damage projections.