LOCA statistical downscaling

2022-11-13: LOCA version 2 (CMIP6) Tmin, Tmax, and Precipitation have been completed for the North American domain at 6 km resolution, covering up to 10 ensemble members from 27 CMIP6 models, with SSPs 245, 370, and 585 included. Data are available for downloading now via globus. You can get a free globus acocunt for downloading data if you are not a commercial user. We are also working on uploading the data to the Green Data Oasis for those who cannot access globus. This is a different data set than the California data, which is at 3 km spatial resolution, created using a hybrid downscaling technique, and includes more variables than the North American data. The California data will be available soon.

2016-09-12: LOCA (CMIP5) data are now available for download from the following sites:

Please read this special note about the LOCA data calendar; LOCA has been interpolated to a standard calendar from the various original global model calendars.

The LOCA elevation field is now available.

2015-06-04: The powerpoint from the LOCA webinars held on June 1st and 4th, 2015 can be found here.

LOCA is a statistical downscaling technique that uses past history to add improved fine-scale detail to global climate models.

We have used LOCA to downscale 32 global climate models from the CMIP5 archive at a 1/16th degree spatial resolution, covering North America from central Mexico through Southern Canada. The historical period is 1950-2005, and there are two future scenarios available: RCP 4.5 and RCP 8.5 over the period 2006-2100 (although some models stop in 2099).

The variables currently available are daily minimum and maximum temperature, and daily precipitation. Over the next year we will begin running the VIC hydrological model with the downscaled data, which will give many more variables, such as snow cover, soil moisture, runoff, and humidity, all at a 1/16th degree spatial resolution on a daily timescale.

From here you can:

When LOCA data is publicly released, you will also be able to access it from here.