Drought Analysis in the US Affiliated Pacific Islands
1. Drought Analysis in the
U.S. Affiliated Pacific
Islands
Richard R. Heim Jr.
NOAA / NESDIS / National Climatic Data Center
Asheville, North Carolina USA
USDM Forum Workshop
Reno, NV, April 2015
2. 2
Overview
• Differences: Nature of drought and drought
monitoring in the U.S. Affiliated Pacific Islands
(USAPI)
• Data monitoring tools
• How can we do it (monitor drought)
• Progress: What we have done so far
• Next steps
3. 3
Nature of Drought and Drought
Monitoring in the USAPI
• Geography
– Large spatial extent
– Isolated small islands
• Varied Populations
– From only a few people
to 175,000 people
4. 4
Nature of Drought and Drought
Monitoring in the USAPI
• Different Hydrology
– Different from CONUS; different High Islands vs. Low Islands
– Groundwater lenses, few reservoirs or streams, no snowpack
storage, rainwater catchment important
No reservoirs or streams.
Rain catchment important.
A few have streams and reservoirs.
SALT WATER
5. 5
Nature of Drought and Drought
Monitoring in the USAPI
• Different Meteorology & Agriculture
– Tropical climate – precip more important than temperature; mean and
standard deviation of precip can be large
• Guam: 165” in 1997 (El Niño year), 55” in 1998 (El Niño + 1 year)
– Minimum amount of rainfall needed each month to meet water needs
– Tropical cyclones
• Can end meteorological &
hydrological drought overnight, but
agricultural drought may continue
for months
• Timing – beginning of wet season =
good; beginning of dry season = bad
(dry kindling/fuel to feed fires)
– Drought can begin rapidly, intensify
rapidly and end rapidly
6. 6
Data Monitoring Tools
• Drought monitoring data very limited
– Daily precipitation data
– Monthly SPI for some (primary) stations with long records
• Other tools lacking
– No Palmer Index
– No VHI (islands too small)
– No usable streamflow or groundwater observations
– No soil moisture observations or modeled data
– Timeliness & availability of data an issue for secondary stations
• Impacts
– Crop damage, wildfires, low streamflow (where streams exist),
water supplies run out
– Water table fluctuates with ENSO status: Low Islands coralline
& porous, High Islands coralline & basaltic & not so porous
7. 7
Drought Monitoring Criteria
• Monthly precipitation trigger: Minimum amount of
rainfall needed each month to meet water needs
– Monthly precipitation > minimum (8” or 4”) No drought
– Monthly precipitation < minimum for 2 or 3 months Drought
• Strong seasonality of precip can result in some months
always > min and other months mostly < min
8. 8
Drought Monitoring Criteria
• Monthly SPI
– Can be useful for determining Dx intensity once drought is
established, but not for triggering drought
• Monthly Percent of Normal Precipitation
– Not as useful if normal is too much different from the monthly
minimum precip drought trigger
9. 9
Drought Monitoring Criteria
• Daily precipitation trigger
– Weekly minimum rainfall needed to meet water needs =
(monthly min) / 4
– 2 or 3 consecutive weeks of no rainfall or low (below weekly
min) rainfall Drought onset
D0 D1
Last 90 days – Dec 15-Mar 15: Last 30 days – Feb 13-Mar 15:
Guam – Feb 2015: 0.18 in.
Driest month on record
10. 10
Drought Monitoring Criteria
• Impacts – Is the lack of rain causing any problems
for the human sectors or the environment?
– Crop damage, wildfires, low streamflow (“lowest I’ve seen in
many years”), water supplies get low or run out, groundwater
becomes brackish
– Useful for determining changes in Dx intensity
– Can also be used as a drought trigger
12. 12
Status: What Have We Done So Far
• Monthly data:
– Provided in conjunction with PEAC conference calls (2nd Thurs.)
– Monthly precipitation, monthly SPI (primary stns), impacts
• Monthly Analysis
– Done manually in spreadsheets
13. 13
Status: What Have We Done So Far
• Monthly mapping:
– Drought condition plotted as points instead of polygons due to
small size of islands
Hawaii, Big Island:
4,027 mi2
93 mi across
Guam:
212 mi2
3.7 mi x 29.8 mi
Kwajalein*:
1.2 mi2
2.4 mi x 0.6 mi
Hawaii Kwajalein
USDM:
15. 15
Status: What Have We Done So Far
• Daily data:
– Obtained from several sources:
• NWS web pages, xmacis web site, automated stn web pages @ WRCC
– Daily precipitation, impacts
16. 16
Status: What Have We Done So Far
• Daily data for weekly analysis:
– Timeliness of daily data can be an issue for the secondary
stations through xmacis On March 20 …
Can do
weekly
analysis
17. 17
Status: What Have We Done So Far
• Daily data for weekly analysis:
– Timeliness of daily data can be an issue for the secondary
stations through xmacis On March 20 …
Not enough
data for
weekly
analysis
18. 18
Status: What Have We Done So Far
• Weekly analysis:
– Done manually in a
spreadsheet
19. 19
Status: What Have We Done So Far
• Weekly mapping:
– Some stations plotted as “No Data” this week,
depending on availability of data
20. 20
Next Steps
• Proof of Concept
– Manual analysis – monthly since Oct 2012 & weekly since Dec 9, 2014
– Proves weekly assessment of USAPI drought conditions can be done
– Establish methodology and data requirements so that they can later be
automated
21. 21
Next Steps
• Operational transmission of daily precip data
– Remote islands will provide daily data to Weather Service Offices via Chatty
Beetle or HF Radio.
– NWS observers will input daily data into WxCoder III system, which Richard
will have access to. COOP stations should be transmitted to NCDC in this
manner.
22. 22
Next Steps
• Automated system for downloading and analyzing daily
precip data
– Ingest the daily data within NIDIS Portal environment
– Each day, compute moving 7-day, 14-day, 30-day precip total; number of
days since X inches (0.10, 0.25, 0.50, 1.00”)
– Each month, compute monthly precip total
– Manual – gathering of data ~60 minutes; analysis ~30-60 minutes
– Automated – gathering of data & calculation of indices will be automatic (0
minutes); get analysis time down to 15-30 minutes or less.
• Get approval from USDM Administrators to make USAPI
operational as part of the USDM
23. 23
Thank You!
• Acknowledgements – This work would not be possible
without:
– NWS Offices and Partners in the USAPI
– PEAC Center (Pacific ENSO Application Climate Center)
– NDMC (National Drought Mitigation Center)
– And probably others whom I’ve forgotten
Groundwater lenses can decrease from top and bottom -- top due to lack of recharge and/or withdrawal through wells; bottom due to salt water intrusion.
Climatology and seasonality of precipitation at a station can determine usefulness of SPI and PNP.
SPI: For wet stations and wet season, SPI can be negative yet actual precip can still be > 8” or 4” minimum. For dry season, SPI can be positive yet actual precip can be well below the monthly minimum.
Percent of Normal Precip: Same dilemma as SPI. PNP can be below normal during wet season while actual precip is above the monthly minimum. PNP can be above normal during dry season while actual precip is well below the monthly minimum.
This is the Drought Severity Classification table from the NDMC web site (http://droughtmonitor.unl.edu/AboutUs/ClassificationScheme.aspx) which is used in the preparation of the U.S. Drought Monitor. It scales (to the USDM Dx percentile categories) some of the drought indicators available for drought monitoring in the 50 States & Puerto Rico. This slide contrasts what is available for drought monitoring on the mainland vs. what is available for drought monitoring in the USAPI.