clim4health is an R package designed to load, transform, process, plot, and save climate data. It is specifically designed for use in downscaling and verification of seasonal forecast data.
The clim4health package is intended for use with other packages developed within the HARMONIZE project, to spatially and temporally harmonize multiple sources of data for use in health impact studies.
This vignette provides an overview of core terms used within the clim4health package, in particular those relating to climate data, downscaling, and verification.
| Term | Definition |
|---|---|
| Bias | The systematic differences (for example, a tendency to overestimate temperature) between a climate model and observed climate. |
| Calibration | The statistical adjustment of raw model outputs to correct for systematic biases and improve reliability using the observational reference. Calibration typically involves comparing historical forecasts with observations to develop correction factors that are then applied to new forecasts. |
| Climate Model | Mathematical model for quantitatively describing, simulating, and analysing the interactions between the atmosphere and underlying surface (e.g., ocean, land, and ice), relying on underlying physical principles and equations. Climate models are used to simulate past, present, and future climate conditions. |
| Climatology | The long-term average of meteorological variables calculated over a specific baseline period. |
| Downscaling | The process of translating coarse-resolution global model forecasts to finer spatial scales to provide more localised predictions. Downscaling can be statistical (using empirical relationships between large-scale and local conditions) or dynamical (using higher-resolution regional models nested within global models). Downscaling may also include bias adjustments to reduce model biases. |
| Ensemble | A collection of multiple forecast simulations (called members) for the same time period, created by running a model with slightly different initial conditions or model parameters. Ensembles help quantify uncertainty by showing the range of possible outcomes rather than a single deterministic prediction. |
| Forecast date | The future date or period for which a prediction is made. |
| Leadtime | The interval between initialisation date and forecast date. |
| Hindcast | Historical forecast initialised using observed conditions. |
| Hindcast period | A standard period (1994 to 2016) used to calibrate and assess the skill of the forecast model. |
| Initialisation (start) date | The specific date when a forecast model begins its simulation. |
| Reanalysis | A dataset created by combining historical observations with a consistent modern forecast model to produce a comprehensive, physically coherent record of past weather conditions. Reanalysis datasets provide a gridded, gap-free representation of historical conditions and are often used as reference data for model development and validation. |
| Skill | The ability of a forecast model to correctly predict an outcome. It can be measured using a variety of different metrics. |
| Uncertainty | The range of possible outcomes in a forecast, typically represented by the spread of ensemble members. Greater uncertainty indicates a wider range of possible outcomes and typically lower confidence in any specific outcome. |
| Verification or validation | The process of evaluating forecast performance by comparing predictions to observed outcomes. Validation metrics may include measures of accuracy, bias, reliability, and skill relative to reference forecasts like climatology. |