The Climate Hazards Center has developed a number of screening processes to quality control inputs and diagnostic tools to evaluate the various aspects of CHIRP and CHIRPS. The various screening routines and resulting products are described in more detail below.

 

Screening and Diagnostics

 

Screening processes are designed to test station inputs to the CHIRPS process, and identify potentially bad data.  There are a number of different quality control standards used to identify suspicious data, and remove it from the processing stream before it goes into the CHIRPS process.  These tests, and any available outputs or results, are explained below.

 

 

 

The daily GTS and GSOD values undergo screening to flag potentially missing values coded with zeros. This is a substantial problem with both of these information sources and can produce inaccurate low values in the midst of a rainy season, as missing data are coded as zeros and passed through the automated GSOD and GTS networks. If a daily GSOD or GTS value is zero for a given day, but the daily CHIRP precipitation estimate is above the long-term (1981-2014) average daily rainfall intensity at that pixel, that daily station value is treated as missing instead of zero.

 

 

Some GSOD data are found to exhibit repeating values of meaningful rainfall (greater than 10mm) on consecutive days. For example, a string of daily observations might be: 0.0000, 0.0000, 87.653, 87.653, 0.0000, .... If a monthly GSOD record has three pairs of repeated values, or the amount of precipitation being duplicated (i.e. the sum of the repeated values) is greater than 30 mm in a month, the monthly record is omitted. A map of these locations is available for every month here (https://data.chc.ucsb.edu/products/CHIRPS-2.0/diagnostics/GSOD_dups_pngs/).

 

 

For each new run of CHIRPS, a number of quality control steps are performed. During this process, maps are made showing the location and number of points excluded due to:

 

Bad Z-score: a Z-score value more extreme than +/- 4.0. 

False Zeros: Station reports a zero, but CHIRP > 7mm for a pentad, or 20mm for a month.

Extreme Values: Station value greater than 2000mm, or greater than 5 times CHIRP (with CHIRP being > 20mm).

 

The monthly maps can be found here:

 

The pentad maps can be found here:

 

In each of these directories, the file names for the products identified above should be self-evident.  Bad Z-score maps are listed as “bad_zscores.YYYY.MM.png”, where YYYY is the 4-digit year, and MM is the 2-digit month.  The false zeros product are listed in the appropriate directory as “false_zeros.YYYY.MM.png”, and the extreme values product is identified as “stn.gt.5xchirp.YYYY.MM.png”.  These graphics identify the number and location of stations which were flagged by the listed quality control test.

 

As part of our quality control effort, when we ingest new sources of station precipitation time series, we compare each new station to other stations previously ingested into our database. We want to weed out stations that are statistically and/or visually different from their neighbors. To this end, we have developed software that calculates correlation and difference statistics to produce an index that we can use to rank the goodness-of-fit with the neighboring stations. We also generate graphics that display the time series of the station with the neighboring stations’ time series.  This allows us to visually inspect stations where the statistics suggest a poor fit with its neighbors during that station’s wettest 3-month period. Individually, we examine the plots and flag those that we suspect may contain erroneous data. We include information about elevation, climatology, geographic locations, and station sources in the plots so that we can determine if the differences we see can be explained by these other variables. See the example plot below.

 

The algorithm selects a station from our precipitation database for a given source and searches for neighbor stations within an increasing radius from the station until a minimum number of "quality" stations is found. This number is usually four, but it can be adjusted if stations are sparse in the area of interest. A quality station has enough observations (5-15 depending on the data set's temporal length) within the time series of interest. For CHIRPS, we are interested in the time period 1981 to present. At each search iteration, if the minimum number of quality stations is not found, the search radius is increased up to a maximum of 150km. If the minimum is not met at the maximum search radius, then the station is skipped and noted in a log file. When enough quality neighbors are found, the distance weighted mean (DWM) of the neighbors is calculated, and the correlation coefficient (R) and the difference between the station and the DWM are calculated. The difference is then divided by the DWM to express that as a fraction of average error (FAE). These values are used to compute a composite index of the overall "badness" of the station compared to the neighbor's DWM. The badness index (BI) is calculated with:

 

 

The BI is calculated for the 3 wettest months at the station location and added together to produce a "total badness index" (TBI) for each station. The time series of the station (blue), and its neighbors (distance-weighted line thickness), is plotted along with the DWM (green) for each of the wettest three months. Each of these graphics contains the latitude and longitude, elevation, country name, and source name of the station. The nearby stations’ distances from the compared station, differences in elevation, station, source IDs, and station names are listed to the right of each time series plot in order of distance to the station. These values are color-coded to match the color of the time series plot line. A plot of the locations of each station selected within a 150 km radius is generated and color coded to the upper-right of each time series plot. The station under examination is plotted in the middle as a triangle, and the neighbors are plotted geographically around it. The climatological precipitation value (CHPclim) for each station is plotted along the right side of each plot as a reference to use in determining station validity. The boundary country the station is located in is plotted in the upper right of the graphic, along with its location and the outline of the station location box for reference. Below this is a colorized elevation map of the searched area to help determine if the topography is affecting the station comparisons. In the lower right corner of the plot, the TBI is printed. The names of these graphics files are prepended with the TBI to allow for numerical sorting by the computer’s file system. Typically, only stations with TBI greater than 100 are saved for viewing, since TBI values below that are stations are in very good agreement with their neighbors.

 

Reality Checks