ACRE Pacific presentation at ICSHMO'12 - Andrew Lorrey (NIWA)

Presentation detailing the ACRE Pacific initiative, related projects, and current work being undertaken on Pacific Island data rescue with the support of the French Pacific Fund.

Transcript

1. ACRE  PACIFIC:  INCEPTION,  CURRENT   WORK  AND  FUTURE  DIRECTIONS     Andrew  LORREY1,  Penehuro  LEFALE2,  Howard  DIAMOND3,  Juerg  LUTEBACHER4,  Petra  CHAPPELL1,     Philip  MALSALE5,  Rob  ALLAN6  and  Pacific  Island  Meteorological  Services   1  Na@onal  Ins@tute  of  Water  and  Atmospheric  Research,  Auckland,  New  Zealand   2  MetService,  Kelburn,  Wellington,  New  Zealand   3  Na@onal  Oceanic  and  Atmospheric  Administra@on,  Global  Climate  Observing  Program,  USA   4  Department  of  Geography,  Justus-­‐Liebig  University,  Giessen,  Germany   5  Vanuatu  Meteorological  Service,  Port  Vila,  Vanuatu   6  UK  Met  Office,  Hadley  Centre,  Exeter,  UK     Special  thanks  to  our  funders  (above),  Davina  Ashford,  Philip  Brohan,  Gil  Compo,  Joelle  Gergis  and  Clive  Wilkinson  
2. Outline   •  Background  on  related  projects  (ICU  &  SPRAT)   •  ICSHMO9  findings  for  SPRAT  (2009)   •  Rob’s  request   •  APN  workshop,  Auckland  (2010)   •  Progress  with  current  work  –  ACRE  Pacific  foci  (2010-­‐Present)   •  Discussion  and  concluding  thoughts  (ACRE  Pacific  guidelines)  
3. ICU  and  SPRAT   •  Island  Climate  Update   Seasonal  rainfall  forecast  for  the  SW  Pacific   Regional  pa[erns  based  on  an  ensemble  of  global  climate  models   Valida]on  using  sta]on-­‐based  observa]ons  provided  by  PINMS     •  South  Pacific  Rainfall  Atlas  (SPRAT)   Project  supported  by  NOAA  (since  2008)   Linked  to  the  Island  Climate  Update,  data  rescue  efforts,  and  regional  weather  and  climate  record  stewardship     Goals:     Assess  regional  data  coverage  for  historic  rainfall  data   Rescue  historic  rainfall  data  (digi]sa]on,  keying,  analysis)   Aggregate  sta]on  results  into  regional  maps  resolved  at  monthly,  seasonal,  and  bi-­‐annual  resolu]on   Produce  an  online  resource  with  commentary  about  causes  of  regional  rainfall  pa[erns  seen  in  the  maps  
4. ICSHMO9  findings  for  SPRAT   1.  100s  of  sta]ons  in  the  SW  Pacific  region   2.  Hundreds  of  thousands  of  monthly  rainfall  entries,  but  many  sta]ons  have   less  than  10  years  data   3.    Fewer  sta]ons  for  smaller  island  na]ons   4.    Significant  coverage  decrease  prior  to  the  early  1950s,      and  during  WWI  and  WWII.   5.    Good  spa]o-­‐temporal  coverage  during  the  mid-­‐1950s  to  1960s   6.    There  is  incomplete  temporal  coverage  for  some  sta]ons    that  cover  1951-­‐2009   7.    Government  transi]ons  and  emergence  to  independent    na]on  status  ofen  see  data  gaps   8.    Manual  to  automated  observa]on  changes  ofen  see  gaps   9.    Splicing  may  be  possible   10.    Common  period  of  overlap  and  most  complete  period  of    coverage  is  1961-­‐1990.  
5. Rob  makes  a  request   Drew,   Following  the  ACRE  WG1  meeGng  in  Bologna  last  week,  Roger   Stone  forwarded  me  the  following  with  links  to  APN  calls  for   proposals.     The  thought  in  Bologna  talking  to  Roger,  was  that  if  we  were   looking  to  try  and  get  a  major  push  on  historical  Pacific  weather   observaGons  underway  one  approach  might  be  to  do  it  with  a   specific  focus  in  mind.  His  thought  was  in  making  sure  that  there  is   enough  data  for  Gil's  reanalyses  to  have  the  best  chance  of   reconstrucGng  a  200  or  so  year  record  of  ENSO  events  and  episodes.   It  should  all  be  interlinked  with  what  KNMI  and  the  Indonesians  are   looking  to  do,  your  NIWA  efforts  linked  to  Australia  (?)  and  any   Chilean  'chapter'  of  ACRE  that  we  can  develop  on  historical  data   digiGsaGon.  Not  sure  if  such  a  project  would  fit  any  of  the  APN  calls   that  Roger  highlighted?     Anyway,  I'd  be  interested  in  your  thoughts  on  this.  Of   course  a  linked  aspect  would  be  to  bring  in  the  Pacific  islanders   themselves,  as  you  indicated  in  the  RMSD13  meeGng  details  e-­‐mail  -­‐   I'd  be  interested  to  hear  how  best  to  engage  them  more  directly  in   our  exisGng  and  ongoing  data  efforts.   Cheers,  Rob.   Lammington  ACRE  workshop  2009   O’Reilly’s  rainforest  retreat   June  2009  we  agreed  to  submit  an  APN  CAPABLE  applica]on  to   support  a  workshop  and  garner  support  for  ACRE  
6. APN  workshop  –  Auckland  2010  
7. Research  Progress   •  SPRAT   •  SPEArTC  and  GRiT  (TC  analysis)   •  Data  rescue  related  work:   – Historical  weather  diaries   – Log  of  Logs  /  Ships  logs   – French  Pacific  Fund   – NIWA  –  Climate  Present  and  Past   – Small  PI  contribu]ons  to  SIRCA     Projects  associated  with  ACRE  Pacific  either  directly  within  current  work  plan  or   through  collabora]ons  with  partners  
8. South  Pacific  Rainfall  Atlas  (SPRAT) Plohng  of  sta]on  rainfall  anomalies  in  Google  Earth  (now  in  version  3)  for     use  by  PINMS  (visual  valida]on  of  ICU  forecasts)  and  climate  researchers  (ENSO   pa[erns)   Reveal  coherent  areas  of  enhanced  &  suppressed  rainfall  (possibly  ENSO-­‐related)   •  Add  value  to  our  understanding  of  atmospheric  circula]on  anomalies  and   impacts  in  the  region   •  Regional  reference  material  in  atlas  format  that  can  be  accessed  by  PINMS   •  Addi]onal  version  for  non-­‐specialists  
9. SPRAT:  viewing  regional  rainfall   anomalies  in  ENSO  phase  space   Lorrey  et  al.,  2012.  ReconstrucGng  the  South  Pacific  Convergence  Zone  posiGon  during  the     pre-­‐satellite  era:  a  La  Niña  case  study  Monthly  Weather  Review,  In  review   1961-­‐90  Climatology  (Nov-­‐Apr)   Anomalies  (%  normal)  for  different  degrees  of  ocean  –  atmosphere  coupling  
10. Southwest  Pacific  Enhanced  Archive   for  Tropical  Cyclones  (SPEArTC)   See  abstract/presenta]on  of  Diamond  et  al.,  this  conference  
11. SPEArTC:  GrIT  analysis  of  TC  sinuosity   rela]ve  to  ENSO  event  coupling   (co-­‐author  Philip  Malsale’s  MSc  thesis  work,  affiliated  with  Vanuatu  Meteorological  Service  and  USP)  
12. Data  rescue:  Weather  diaries   • Twice daily T, P, wind dir. • 1839-1844; 1848-1851 = • More than 13000 obs. • PLUS • Comments section include: Frost, hail, wind strength, sunsets, floods, ice, cloud, fog, relative rainfall, drought, thunder, lightning, and bio indicators • Comments section examined on first trip • SNOW • Dirty weather
13. •  8 July 1850 – 9am temp 1.7°C; severe frost; not an outlier for the month! •  Modern mean daily min T in July is ~8°C •  Davis diary temperatures (9am) compared to Kaikohe AWS (Mean Daily T as a proxy for 9am). Similar range of interannual variability signal observed through the year, but summer and winter possibly cooler in mid 1800s Data  rescue:  Weather  diaries   •  Outliers and extreme pressures below 25th percentile may indicate extropical cyclones passing close by •  October 1842 – early start to TC season? And a Direct Hit??!!?? •  Proximity of ex-tropical cyclone or Tasman low (ECL or sub-polar origin) indicated by remarks? Temperature                             Pressure  
14. Frost, snow and cold in “the Winterless North” "To  MR.  CHARLES  CROCKER,     Stourton  Caundle."     "July  30,  1849.—The  hills  were  covered  with   snow,  the  first  ever  seen  by  the  naGves   inhabiGng  this  part  of  New  Zealand.  The   Putahi  was  also  covered.”   “July  31.—This  morning  the  hills  were  again   white  with  snow.”     Davis  weather  diary  notes  ICE  ¼  inch  thick  on  the  surface  of  a  duck  pond  on  the   property  in  Waimate  North,  15  July  1839  (close  to  Kaikohe).  Temp  readings  at   9AM  were  as  low  as  4.4°C  on  that  day.    (Also  note  it  was  as  low  as     1.7°C  in  1850  at  Kaikohe)   Excerpt  from  Coleman,  J.N.  1865.     •  Frost (1985-2004) climatology at Kaikohe •  Max days occurrence of ground frost = 2 days; Max days occurrence of screen frost = 1 day; None in last two years •  Davis diary 1848-1851: 82 days where comment of ‘frost’ occurs Data  rescue:  Weather  diaries  
15. Data rescue: Ships’ logs Ships’ log books may contain past weather data and observations useful for reconstructions –  Pressure –  Temperature –  Rainfall –  Storm/tropical cyclone events –  Wind –  Anecdotal information (wave climate, extreme events, icebergs) If ships were anchored in harbours for extended periods, logs may contain longer-term records useful for climate reconstruction. E.g. NZ for Maori Land Wars (2-3 years); convict ships anchored in Australian harbours (makeshift prisons) for a number of years. H  M  S  North  Star  destroying  Pomare's  pā,  Northland  New   Zealand  1845.  Pain@ng  by  John  Williams   View  from  close  to  the  pā  site  into  the  Bay  of  Islands  
16. Data Rescue: The Log of Logs •  Compiled by Ian Nicholson –  Ex Commodore of Royal Australian Navy –  Researcher of nautical records •  “A catalogue of logs, journals, shipboard diaries, letters, and all forms of narratives, 1788-1998, for Australia and New Zealand, and surrounding oceans.” •  Three volumes: published 1990, 1993, 1999 •  Sources from around the world (mainly concentrated in NZ, Australia, UK, USA and Europe) –  Museums, libraries, universities etc The Astrolabe. Source: Log of Logs Source: Log of Logs
17. Log of Logs analysis •  Ships that visited NZ and SE Australian waters prior to 1900 •  Keyword search through PDFs of LoL volumes –  E.g. Wellington, Auckland, Lyttelton etc –  E.g. Sydney, Hobart, Melbourne etc •  Only recorded ships that had a record of >1 year (except pre-1820, records of <1 year were included because of less ships at this time) –  Wanted to target potential for longer-term weather data •  Recorded ship’s name, year(s), where it went on its voyage, and sources of information.
18. •  OCR for all three LoL volumes •  Key word searches conducted •  509 separate ships across all three volumes, that came into NZ waters between 1787 and 1900. –  Many had multiple voyages •  Peak around 1870s-1880s •  Many of the ships with longer records (>5 years) are migrant ships that did multiple voyages from UK to NZ •  Many different sources of information across the world (museums, libraries etc) (mainly NZ, Australia, UK, some USA) –  Alexander Turnbull Library, Hocken Library, NZ National Archives etc –  Australia Joint Copy Project (AJCP), Sydney Maritime Museum, British Museum –  Most ships went to Auckland, Bay of Islands, Wellington, Lyttelton and Dunedin Number of ships in NZ waters 1787-1900 0 20 40 60 80 100 120 1780 1800 1820 1840 1860 1880 1900 Year Number of ships •  Provided  a  similar  analysis  to  the     SouthEast  Australia  Recent  Climate  History  (SEARCH)  project,     based  at  University  of  Melbourne  for  SE  Australia.   Major  harbour  posi?ons  in  New  Zealand.   Source:  World  Port  Source  (2009)  
19. Data  Rescue:  French  Pacific  Fund     Loca]ng  data  for  New  Caledonia  and  French  Polynesia Louis  Le  Breton  (Paris,  1818-­‐1866)  L’Astrolabe  and  La  Zelee  in  the  Torres  Strait,  1843   Peabody  Essex  Museum,  Massachuse[s  USA   Examples  of  culturally  significant  items  gathered  from  different    SW  Pacific  islands  on  early  voyages   Peabody  Essex  Museum,  Massachuse[s  USA   0   5   10   15   20   25   30   35   40   45   50   1780   1800   1820   1840   1860   1880   1900   All  of  SW  Pacific   0   5   10   15   20   1780   1800   1820   1840   1860   1880   1900   French  Polynesia  &  New  Caledonia   combined  
20. 0   20   40   60   80   100   120   140   160   1786   1791   1796   1801   1806   1811   1816   1821   1826   1831   1836   1841   1846   1851   1856   1861   1866   1871   1876   1881   1886   LoL   Clive   Overlap   0%   20%   40%   60%   80%   100%   1786   1792   1798   1804   1810   1816   1822   1828   1834   1840   1846   1852   1858   1864   1870   1876   1882   1888   LoL   Clive   Overlap   Poten]al  resources   Data  Rescue:  French  Pacific  Fund     Loca]ng  data  for  New  Caledonia  and  French  Polynesia Monthly  pressure  –  Tahi],  1845   Thousands  of  sources  are  poten]ally  available;  independent  searches  ofen  turn   up  different  results  with  limited  overlaps.  Important  to  work  together  and   con]nue  support  for  historical  research  to  iden]fy  data  sources  
21. NZ  MSI:  Climate  Present  and  Past   •  Rescue  of  10  NZ  sta]ons  iden]fied  for  inclusion  in   ISPD  and  extended  reanalysis  without  radiosondes   effort  (SIRCA)   •  38000  measurements  gathered  for  Auckland  over  6   sites  for  1853-­‐1958;  Dunedin  is  next  on  the  list  
22. Small  island  contribu]ons  to  the   extended  reanalysis  via  ACRE  Pacific   •  Improving  spa]al  and  temporal  density  of  surface  pressure  for  the  Southwest   Pacific  back  into  the  1800s   •  Cook  Islands  Meteorological  Service  (4  sta]ons,  2  NCI/2  SCI  back  to  early  1900s) •  Niue  (Alofi  sta]on,  back  to  early  1900s)   2005-2009: 800+ station years digitised with support from NOAA and the Ministry for the Environment, NZ -  Cook Islands (Rarotonga, Penryhn) -  Kiribati (Christmas Is, Tarawa) -Tuvalu (Funafuti) -  Tokelau -  Samoa (Apia) -  Niue (Alofi)
23. SPRAT  +  SPEArTC  +  20CR  =   Past  SPCZ  behaviour   Lorrey  et  al.,  2012.  ReconstrucGng  the  South  Pacific  Convergence  Zone  posiGon  during  the     pre-­‐satellite  era:  a  La  Niña  case  study.  Monthly  Weather  Review,  In  review  
24. Discussion   The  SW  Pacific  is  a  region  that  is  cri]cal  for  us  to  examine  if  we  wish   to  understand  how  Earth’s  climate  system  works.       Much  of  what  we  need  to  know  is  s]ll     unwri[en.     We  have  opportuni]es  to  expand  our     knowledge  using  historical  climate  data.     New  visualiza]on  tools  can  bring  these     data  to  life.     The  extended  reanalysis  without     radiosondes  effort  (20CR  and  SIRCA)   provides  a  unique  way  to  integrate  data,       and  gives  us  circula]on  reconstruc]ons  (past  climate  and  weather)  that  can  be   usefully  paired  with  diverse  observa]onal  data  (in  situ  sta]on  data,  shipboard   measurements,  and  tradi]onal  knowledge).     The  veracity  of  the  20CR  and  SIRCA  reconstruc]ons  are  dependent  on  the   temporal  and  spa]al  density  of  observa]ons.  We  need  to  help  by  providing   more  data  to  those  projects  via  ACRE.     Suva  Book,  Fiji  Met  Service  
25. Concluding  thoughts  (from  APN  report)