Introduction

Advantages of a Probabilistic Approach

Statistical Models

Large-Scale Predictors

Fitting Statistical Models - Cross Validation

Data and Data Quality Control

Data Sources

Hour of Observation

Precipitation Biases

Interpolating to a Grid

Covariability between Grid Points and Variables

Climate Models and De-biasing

Performance

Future Projections

Home

Downscaled Climate Projections

Data and Data Quality Control

Precipitation Biases

Many COOP stations report days with weak precipitation as zero precipitation leading to an under-reporting of "wet days" and an artificial precipitation amount histogram. These same stations also tend to round precipitation to the nearest 5 hundredths of an inch instead of the nearest hundredth of an inch (see Daly et al. 2007). For the CMIP5 downscaling, poor stations were identified by calculating the ratio of days with precipitation amounts from 1 to 4 hundredths of an inch to the days with 5 hundredths of an inch. The poor stations are then discarded from any further calculations. For the new CMIP6 downscaling, a new method for correcting the bias that keeps the station is used.

---

Home