Climate Data Rescue for the Hawaii

1
Climate Data Rescue for the Hawaii

• Pao-Shin Chu, Sean Newsome, Ying R. Chen
• Hawaii State Climate Office
• Department of Meteorology
• University of Hawaii
• Honolulu, HI
• To be presented at CDMP Workshop, Nov 4-5, 2009

2

• Comprehensive and systematic collection of daily
  rainfall data became available since 1905
• Data sources include sugar plantations, pineapple
  growers, ranchers, irrigation companies, Honolulu
  Board of Water Supply, NWS, USGS, and private
  citizens
• Over 1,900 rain gages with different record
  lengths from six Hawaiian Islands (Kauai, Oahu,
  Maui, Lanai, Molokai, and Hawaii)
• Most data is either in the original reports (hand
  written or computer printout) or in the published
  documents

3

• State Key Number (SKN) for each rain gage was
  created in 1950s but many gages have been
  discontinued or added. Some gage have more than
  one SKN or no number at all.
• Once inventoried, quality checked, filled, and
  digitized, these data sets will allow studies of
  long-term climate changes in rainfall and extreme
  events in the Pacific Ocean. Reliable daily
  rainfall records for other tropical Pacific
  islands are relatively short or sparse.

4

• Proposed Tasks
• Data inventory and updating (compile and
  catalog all historical daily rainfall records
  identify SKN for each gage, label manually SKN on
  each sheet to facilitate keying process)
• Metadata (identify latitude, longitude,
  elevation, station name, observer name, SKN, data
  length, and data format in spreadsheets use
  ArcGIS DEM to determine gage elevation if
  necessary note any relocation and change in
  observers or instruments)
• Data transmission (handle, pack, and ship data
  in increment to a contractor for scanning and
  digitizing)

5

• Fig. 1. Computer recorded daily records
  for January 1989, from the Kekaha Sugar Company,
  Kauai. State Key Numbers are listed at the top of
  each column, site names are located below. Each
  number in the entry indicates a daily reading and
  monthly totals in inches are located two rows
  from the bottom.

6
No state key number!
No observation during the weekend.

• Fig. 2. Daily rainfall records for
  January 1989 from Lihue Sugar Plantation. Missing
  records such as state key number and observer
  have to be identified prior to the keying
  process. Identified state key numbers are
  manually labeled in red above each site. Also
  note that state key number for one gage (M-8) has
  not been assigned. No footnotes or key are
  available to determine the meaning of the dash
  line used but it looks like no observations for
  the weekend for some gages.

7
Trace

• Fig. 3. An unknown gage in Hawaii for
  June 1988. Partial daily rainfall and
  temperature readings are given. It is
  undistinguishable if only days when rain fell the
  gage was checked. Blank entry under the rainfall
  column probably indicates zero rainfall value
  because there were readings for temperatures for
  that day. There was no reading for the first six
  days of the month.