Object-based analysis of fronts and frontal precipitation in climate datasets
Extratropical cyclone systems and their associated fronts are essential to midlatitude weather. This work develops and tests an object-based methodology to track fronts and their precipitation in climate datasets. The open-source TempestExtremes software package is used to track frontal objects within the European Centre for Medium-Range Weather Forecasts reanalysis 5 (ERA5) dataset. These frontal objects are co-located with precipitation objects to form a climatology of fronts and frontal precipitation. Frontal precipitation is found to account for at least 50% of precipitation across the midlatitudes, and upwards of 90% in the storm track regions. The National Weather Service (NWS) surface analysis dataset of manually curated fronts is used to compare against the TempestExtremes tracked fronts in ERA5 across North America. Using the NWS dataset, different variations of hit rate and false alarm rate metrics are presented as validation of the objectively tracked framework. Sensitivity analysis is performed for parameters used in the objective tracking, including minimum size, minimum timesteps, and extratropical cyclone co-location distance. This climatology is the crucial first step in validating how climate model simulations represent fronts and their precipitation, which will create greater confidence in future projections of fronts and their precipitation.