CHIRPS FAQ

From CHG-Wiki
Revision as of 10:46, 12 February 2015 by Pete (talk | contribs)
Jump to navigationJump to search

Is there a CHIRPS website?

Yes: http://chg.geog.ucsb.edu/data/chirps/index.html

Where can I get CHIRPS data?

Via a web browser

ftp://chg-ftpout.geog.ucsb.edu/pub/org/chg/products/CHIRPS-1.8/ ftp://chg-ftpout.geog.ucsb.edu/pub/org/chg/products/CHIRPS-2.0/


Via anonymous ftp

ftp chg-ftpout.geog.ucsb.edu
username: anonymous 
password: your_email_address
cd  /pub/org/chg/products/CHIRPS-2.0/

What is the spatial resolution of CHIRPS?

  • Chirps is produced at 0.05 x 0.05 degree spatial resolution.
  • For daily Africa data we also provide a 0.25 x 0.25 degree spatial resolution product.

What are the spatial domain available for CHIRPS?

  • Global 50N-50S, 180W-180E @0.05 deg resolution = maps 7200 x 2000 pixels
  • Africa 40N-40S, 20W- 55E @0.05 deg resolution = maps 1500 x 1600 pixels
  • CAmer-Carib 23.5N-6N, 93W- 57W @0.05 deg resolution = maps 720 x 350 pixels
  • WHem 50N-50S, 125W- 34W @0.05 deg resolution = maps 1820 x 2000 pixels

What are the time steps available for CHIRPS?

How does CHIRPS define a pentad?

  • 6 pentads = 1 calendar month. Each of first 5 pentads in a month have 5 days. The last pentad contains all the days from the 26th to the end of the month.

How does CHIRPS define a dekad?

  • A dekad = sum of 2 pentads. There are 3 dekads in a calendar month.

What are the file formats available for CHIRPS?

  • .tif
  • .bil

What units is CHIRPS in?

Total mm for given time step, mm/pentad, mm/month, etc.

What value is used for Bad Values?

-9999

What is the latency of the CHIRPS product?

  • There is a rapid (gts-only) CHIRPS available 2 days after the end of a pentad.
  • Final CHIRPS (all station data) is available sometime after the 15th of the following month.

What is the temporal range for CHIRPS?

CHIRPS is available from 1981 to near-real time.

Who do I contact with CHIRPS questions?

Pete Peterson

pete@geog.ucsb.edu


What is the spatial density of station data used in CHIRPS?

The closest thing we have available is what we call "reality checks" we post in EWX (Early Warning Explorer).

You can access EWX here.

  • On the left open up the Africa tab.
  • Then double click on RCHECKS-STNS
  • You will have to zoom in to see squares with black dots in the middle.
    • These are 10x10 pixel boxes of the station value on top of the CHIRPS field, a "reality check" to see if the stations match the background field.
    • You can zoom in on a location, then cycle through month/years to see how the number of stations changes with time. It is pretty significant in some areas.

Also, ftp://chg-ftpout.geog.ucsb.edu/pub/org/chg/products/CHIRPS-latest/qc/stations_used/monthly/:

  • Each file represents a single month of CHIRPS for the Globe.
  • There are 5 columns
    • year
    • month
    • latitude
    • longitude
    • station_used?
  • Each file has 52,743 lines representing our total list of anchor stations used in CHIRPS.
  • The last column, station_used is
    • 0 if no station data used for that location/time
    • 1 if we had station data for that location/time.

How was CHIRPS funded?

CHIRPS has been supported by funding from:

  • the U.S. Agency for International Development (USAID)
  • the USAID Famine Early Warning Systems Network (FEWS NET)
  • the National Aeronautics and Space Administration (NASA) and
  • the National Oceanic and Atmospheric Administration (NOAA)

Who uses CHIRPS?

How good is CHIRPS?

For its intended uses, Seasonal Drought monitoring and Trend Analysis, it is very good; see validation-pptx (under construction) For other uses it depends on the questions you are asking.

How do I reference CHIRPS in publications?

Funk, C.C., Peterson, P.J., Landsfeld, M.F., Pedreros, D.H., Verdin, J.P., Rowland, J.D., Romero, B.E., Husak, G.J., Michaelsen, J.C., and Verdin, A.P., 2014, A quasi-global precipitation time series for drought monitoring: U.S. Geological Survey Data Series 832, 4 p. http://pubs.usgs.gov/ds/832/

Is there a copy of the paper online?

Yes.

What's on the agenda for future CHIRPS?

  • redo the CHPclim to include more station data and avoid tiling artifacts
  • include fGSOD data from 1981-1999
  • include fGTS data from 1984-1994
  • include Peru-ANA station data
  • include an additive step to the blending
  • consider re-prioritorizing fGTS above fGSOD
  • consider providing daily CHIRPS for more than Africa


What are some of the changes in version 2.0?

  • Since v1.8 we have added many new stations around the world. In particular, 400 in Peru, 1,200 across Africa from Sharon Nicholson, 11,000 in Brazil, over 1,000 in Mexico (Conagua), 76 from SASSCAL, 500 in Russia, 250 in Central Asia and 400 across Central America. In addition we have topped off data from several African countries. The Conagua stations are available in near-real time and will directly improve our preliminary CHIRPS product.
  • Daily data now available for the Globe
  • NetCDF format added
  • Replaced erroneous stations in Tanzania
  • Expanded diagnostics, stations by Country plots, EWX rchecks, station comparison/*badness* plots, station density maps, pentad and monthly png's of excluded stations,

What are some of the changes in version 1.8?

  • Zambia was a mess in v1.7 due to accidentally ingesting temperature data as rainfall in our database. This has been fixed.
  • Pixels near the coastline would alternate between data and bad value over time in v1.7. This has been fixed.
  • Reran CHIRP from May 2013 through October 2013 to fix problem.
  • Deleted 251 bad stations from our database.
  • Fixed another 660 stations with incorrect lat, long and/or country information.
  • Separated Sudan into Sudan and South Sudan stations.
  • Reduced the number of anchor stations used to merge station data from ~91k to ~50k. The reduced stations had little/no data.
  • Added thousands of stations in Colombia, Brazil, Panama and India.
  • Re-factored and re-organized entire code base to be more streamline and less error prone.
  • Include browse images for Africa