Monday, December 19, 2016

Development and analysis of a homogeneous long-term precipitation network (1850-2015) and assessment of historic droughts for the island of Ireland.

Following up on last week’s blog I would like to present a short overview of my PhD thesis. The overarching aim of my research was to rescue and transcribe (key into excel) hard copy long-term monthly precipitation records for the island of Ireland (see Figure 1). To quality check and assess the long-term precipitation records for variability and change and analyse the records to identify past drought events. In addition, my research aimed to integrate past documentary evidence into the analysis to add confidence to the data and present some of the social and economic impacts from past drought events. Finally, I aimed to reconstruct long-term river flow records utilising the good quality monthly precipitation records and asses the flow for past drought events.
Figure 1. Hard copies of rainfall records held in Met Éireann archives (Photos taken by S.Noone, 2012)

 Summary of key findings:
This research produced a quality assured Island of Ireland precipitation (IIP) network of 25 stations dating back to 1850. The results of the analysis show that the years 1891 and 1964 stand out as the driest winters at nine and six of the IIP stations respectively. The wettest ranked winters across 12 stations occurred in 1877, 1994 and 1995. The summer of 1995 was the driest at 6 stations (east and southeast) while 1976 was driest at 3 stations (midlands and northeast) since 1850. 1861 ranks as the wettest summer for 8 stations located along the west coast while 1958 is wettest for stations in the east. The 2000s also stand out because of wet summers (in 2007, 2008 and 2009). Spring 1947 was the wettest for 15 stations with both 1995 and 1976 notable as the driest springs.
The Mann Kendall trend test results indicate positive trends in winter precipitation and negative trends in summer over the period 1850-2010 (see Figure 2 and 3). The trend results following data homogenisation showed changes in magnitude and direction in trends at some stations. Malin Head has been analysed in previous studies (e.g. McElwain and Sweeney, 2007) and significant annual increasing trends were detected 1890-2003. However, post homogenisation no annual trends were present at Malin Head, with similar results found for winter. In addition, summer trends pre-homogenisation at Malin Head indicates no trend while post homogenisation trends show significant decreasing trends.
These results show the importance of assuring that climate records are homogenous as misleading trends can be present. The trends in shorter records commencing post 1940 are not representative of the detected trends since 1850. The results show that in most cases trends over the period 1940-2010 contradict the trends detected over the period 1850-2015, highlighting the importance of long-term records.
Figure 2 Homogenised winter time series for all stations smoothed with an 11-year moving average (black line). MK Z scores are shown before and after homogenization where applicable (red: unhomogenised; blue: homogenised/no breaks detected) calculated for varying start years (which are given by the x-coordinate). The grey lines indicate ±1.96; absolute values exceeding these bounds are interpreted as significant at the 0.05 level.
Figure 3 same as Figure 2 only for summer.

This research also produced a 250-year detailed historical drought catalogue for Ireland and integrated qualitative historical documentary evidence. The results identified seven major drought rich periods in the IIP station network during 1850-2015 with drought events lasting (>18 months) impacting simultaneously at least 40% of the study sites in 1854-1860, 1884-1896, 1904-1912, 1921-1924, 1932-1935, 1952-1954 and 1969-1977 (see Figure 4). 

Figure 4 Drought signatures showing SPI-12 values for all stations in the Island of Ireland Precipitation (IIP) network for the 7 drought periods identified with island wide impact. Negative SPI-12 values are colour coded according to severity thresholds to highlight periods of moderate to extreme drought conditions.
Results for an extended precipitation series (1765-1849) identified a further seven long duration droughts (>18 months) during 1784-1786, 1800-1804, 1805-1806, 1807-1809 1813-1815, 1826-1827 and 1838-1939. Many of these drought events occurred during or immediately prior to Irish famine events, most notably the Great Irish Famine 1845-1849 (Ó Gráda, 2015).  
An online drought mapping application which highlights three of the droughts identified and their impacts, the map also provides a summary of all droughts and can be accessed at:
Documentary evidence has provided important insights into the impacts from past severe drought in Ireland while highlighted interesting societal responses (See Figure 5 and 6). Impacts include reduced or failed crop yields, increased crop and dairy prices, human and livestock  health  issues,  water  restrictions,  low  reservoir  levels,  water  supply  failures  and hydro-power  reductions. The  work  shows  the  importance  of  combining  qualitative  and  quantitative evidence  of  historical  droughts,  which  provides  crucial  information  allowing  for  a  much clearer understanding of drought development and impacts. 

Figure 5 Letter to the Irish Times published 16th September 1893 proposing exploding dynamite over Dublin to try and induce rainfall.

Figure 6 Circular from the Bishop of Meath authorising the prayer for rain, published in the Irish Times on 2nd July 1887.

 My thesis produced (for the first time) a homogenised precipitation network of 25 stations for the island of Ireland. The precipitation network analysis has contributed considerably to knowledge by providing important insights into variability and change over the longer term. In addition, this thesis has produced (for the first time) a detailed 250-year drought catalogue for the island of Ireland. This work contributes significant new knowledge that can be used for stress-testing the resilience of planned Irish water resource developments. By combining qualitative and quantitative evidence of historical droughts this thesis has provided a more coherent understanding of drought development and historic impacts. Finally, several peer reviewed journals have also stemmed from my research (Murphy et al., 2016; Noone et al., 2015; Noone et al., 2016; Wilby et al., 2015).

McElwain L, Sweeney J. 2007. Key Meteorological Indicators of Climate Change in Ireland. Environmental Research Centre Report available online from: .
Murphy C, Noone S, Duffy C, Broderick C, Matthews T, Wilby RL. 2016. Irish droughts in newspaper archives: Rediscovering forgotten hazards? Weather. Accepted.
Noone S, Murphy C, Coll J, Matthews T, Mullan D, Wilby RL, Walsh S. 2015. Homogenization and analysis of an expanded long-term monthly rainfall network for the Island of Ireland (1850–2010). Int. J. Climatol. doi: 10.1002/joc.4522.
Noone S, Broderick  C,  Duffy C, Matthews T,  Wilby RL, Murphy C. 2016. A 250 year drought catalogue for the island of Ireland (1765-2015) Int. J. Climatol. Accepted
O’ Gráda C. 2015. Famine in Ireland, 1300-1900, UCD Centre For Economic Research Working Paper Series 2015, UCD School Of Economics, University College Dublin. Belfield, Dublin 4.

Wilby RL, Noone S, Murphy C, Matthews T, Harrigan S, Broderick C. 2015. An evaluation of persistent meteorological drought using a homogeneous Island of Ireland precipitation network. Int. J. Climatol. doi:10.1002/joc.4523

No comments:

Post a Comment

Please note all comments are moderated. Comments containing profanities, unwarranted accusations, deemed off-topic etc. shall be deleted. There may be a substantial delay between posting a comment and its acceptance owing to moderator availability.