On the banks of the Shannon at Clonbunny, just south of Athlone, there is an abandoned, semi-derelict farmhouse with the river flowing gently by just a short distance away. The house has partly collapsed, following the flooding that occurred in the area in June of last year.
This was home to the Nally family for more than 100 years. Fintan Nally was born there in 1944, and remembers moving out around 1950 to live in Athlone. The family continued to farm the land and returned to the house every summer without incident for several years afterwards.
Fintan recalls that although winter floods sometimes lapped the doorstep, the house stayed dry until December 1954, when it flooded to a depth of "about a foot".
He says that summer floods were unheard of in those years, and that the floods of that December were exceptional – the River Tolka in Dublin flooded at the same time during prolonged heavy rainfall.
Fintan believes that the Shannon navigation was left in good condition by the British when Ireland achieved independence, but that it has been allowed to deteriorate to the point where summer floods are now becoming the norm and are as bad as the worst winter floods from times past.
This summer has been a welcome exception. But then, the summer of 2013 will be remembered for the good weather and cracking hurling and football championships in the same way as 1976 is talked of by those old enough to remember it.
This year apart, Michael Silke of the IFA agrees with the sentiments expressed by Fintan. Michael farms on the callows or Shannon wetlands some 20 miles south of Clonbunny, close to Banagher, Co Offaly.
He is passionate about the river and the low-lying farmland on its banks, and he recognises the unique habitat that results from being located on a flood plain.
"We never minded winter flooding," he says, "it goes with the territory and we can work around it. We value the habitat and the species that is supports, particularly the corncrake, and we've worked hard in recent decades to try to ensure their survival. But last year there were no corncrakes, we lost our silage crop, and it's not our fault."
Michael claims that Waterways Ireland and the OPW are not doing enough to maintain the channel and to manage river levels when heavy rain is forecast. He has concerns about the severe flooding that occurred in June 2012, when water levels rose rapidly following 100mm of rainfall over a couple of days.
When the river burst its banks it rose to the level of the windows in the old Nally house in Clonbunny, destroying thousands of acres of silage crop and subsequent grazing.
Those floods are just a memory now, but should the local authorities and the bodies in charge of water levels on the Shannon be using the respite this summer has provided to be planning and acting to prevent floods such as those that wreaked such devastation last year?
Both Fintan Nally and Michael Silke believe that there were two reasons for the flooding in June 2012. Waterways Ireland's failure to act quickly to raise the 'boards' on the top of the weir at Meelick when the heavy rains were forecast, and the decision taken in 1979 to maintain the level of Lough Ree at 600mm above previous levels in order to improve navigation.
Brian D'Arcy disagrees. A senior engineer who has worked on the Shannon for most of his life, first for the OPW and latterly for Waterways Ireland, Brian is adamant that Waterways Ireland not only fulfilled its remit, but that there was nothing more that they could have done, given that what happened last summer was effectively a flood of winter magnitude.
"Once the flood waters have 'topped' the weir at Meelick there is nothing that anyone can do about it," Brian maintains.
"At that point, removing weir-boards makes no difference, because there is more water in the river than that channel can take, and the only way that it can go is sideways, across the natural flood plain, the callows."
Last year, he says, the June level of the water was more than 600mm above the top of the weir in Athlone.
Brian also points out that the Rivers Suck and Brosna, which join the Shannon above the Meelick Weir, dump between 40pc and 60pc of the volume of the Shannon flow into the channel during flood periods, effectively acting as a 'choke' that further holds back the flow of water in the main channel.
This year's fine summer may have lulled us into a false sense of security about the Shannon. Was last year a unique event, or are changing weather patterns going to deliver this '100-year flood' on a more frequent basis?
Farmland in the callows was dry this year and crops will be saved without difficulty, but will that be the norm in future?
As the June floods subsided last year I drove around the area to examine the damage. Farmland just south of Athlone was under 100mm of water, even though the worst flooding had receded. Silage crops were beaten flat and rotting in the water, and farmers were feeding animals indoors. On Michael Silke's land in south Galway the summer meadows were waterlogged and rotting.
Water was still thundering across the weir wall at Athlone with frightening velocity, and similarly downstream at Meelick. South of Lough Derg, at the headrace of the Ardnacrusha power station, levels appeared normal and there was no flooding. The problems were all upstream, in the callows.
There is a local view that ESB stores water in Lough Allen and Lough Ree as generation reserve, and that this means that flash flooding is forced across the flood plain. Not so, says Tom Browne of the ESB.
Tom describes this assertion as "misinformation", which he says is tied into the history of the problem.
Ardnacrusha was built in the 1920s and was then the largest generating station in the country. Until the advent of larger stations and the construction of Turlough Hill pumped storage station in the 1970s, there would have sometimes been excessive draw-off of water to maintain generation, particularly in the war years when Ardnacrusha was often the only station operating. The focus changed in the early 1970s to one of managing the river to meet the needs of all stakeholders.
Tom Browne points out that ESB has no interest in Lough Allen and Lough Ree for generation, but the company has an obligation to manage levels in the lakes to ensure dam safety, to provide adequate water supply for fisheries and public bodies, and to minimise flooding. Athlone weir is 123 feet above the ordnance datum at Poolbeg, and the 1934 Electricity Supply Act allowed ESB to draw water down to 121 feet to facilitate continuity of electricity supply in drier months.
In the early 1970s, levels were gradually restored to the 123 foot level to facilitate boat traffic, and in 1979 the company agreed formally with the OPW to maintain 123 feet as a summer minimum. This restoration of the level to what it was in the years before Ardnacrusha was built is often quoted as the reason for summer flooding, although there is no suggestion that summer flooding was an issue when the Shannon was a commercial waterway with related higher levels in Lough Ree.
ESB nowadays allows the flooding level at Athlone to drop by 300mm in mid August and by a further 300mm in mid October as part of its flood management role, and also collects and maintains records of rainfall, water levels and flow data on the river. The company has no responsibility to supply flood warnings, but does issue twice-weekly lake-level forecasts to all relevant stakeholders.
Most observers agree that the problem centres on the river channel between Athlone and Lough Derg. The river here lacks gradient and flood waters simply spread across the flood plain when the levels rise. However, there is also widespread belief that not enough is being done to maintain this channel, and that authorities are slow to respond when flooding is imminent.
Michael Silke showed me a section of the river north of Shaughnessy's Bridge in Co Offaly, pointing out a large silt bank that appears to block around one third of the channel. Michael believes that this bank is obstructing the flow, but it cannot be removed because the National Parks and Wildlife Service (NPWS) will not allow it to be spread on the banks in this special area of conservation (SAC).
The cost of moving the silt away from the area would be prohibitive, but local farmers suggest that a temporary derogation should be sought for this section of riverbank to allow for this remedial work.
Michael Collins of the OPW admits that taking out silt banks and maintaining the channel will give benefits during small flood events, but he says that these will be negligible and will, if anything, only benefit 'the last place to flood' – ie, the highest parts of the flood plain.
He says that summer floods, such as happened last year, are not preventable without major engineering work, costing 'hundreds of millions of euro'.
Given the gradient of the river, he says that such work would have to include the construction of embankments – just deepening and widening the channel won't solve the problem; the gradient is simply too flat.
The OPW currently has no responsibility for the maintenance of the main river channel; they are only tasked with maintaining channels that were altered by them in the course of arterial drainage projects.
They do, however, maintain a number of gauges on the river to record data, and Michael quotes figures showing a slight shift in recent years from late summer to early summer flooding. He acknowledges that this might make it more difficult for farmers to cut meadows, given the late-cutting required to protect wildlife.
However, on the topic of whether the flooding is getting worse, he believes that any such signal within the data – if it exists – is tiny and insignificant.
With regards to the difficulties in removing silt in SACs, his experience has been that the NPWS is as flexible as they can be, and that they usually take a practical approach to sorting out problems.
Michael Collins believes that existing working arrangements between the various bodies are excellent and that the Shannon Catchment Flood Risk Assessment and Management (CFRAM) study, currently being undertaken on behalf of the OPW, will show all the players how best to work together to address contentious issues.
The Shannon CFRAM study is designed to examine flood risk and to see how that risk can be best managed.
When the first part of that study has been completed later this year, the nature of the problem should be clearer, and the completed study – due in 2015 – should provide a template for best practice in the management of the waterway.
It won't solve the problem though; farming on flood plains carries an inherent risk of flooding. That is accepted by the farmers along the river; they are just not convinced that everything possible is being done to minimise the risks in summer time.
Despite the absence of a single authority charged with managing the river, and the risk that this may result in some 'grey areas' where responsibility is not well defined, the various bodies appear to be following their individual remits and doing exactly what they are tasked to do.
Last year's June flood occurred when winter levels of rain were dumped on land saturated following prolonged rainfall; the question being asked by farmers is whether enough was done when the flooding was imminent.
We won't know a full picture until the CFRAM report is completed in 2015, but there is some good news for farmers.
Following a meeting of all parties with Minister Brian Hayes in September last year, it was agreed that calibration tests at Meelick weir would be fast-tracked within the CFRAM project, with these tests carried out as soon as conditions allow, at low water ahead of an anticipated flood.
Weir boards will be lifted and flows measured, to see whether such action could help alleviate flooding in the future. It's an important concession, and will clear up any disagreement around the value of such action when flooding is imminent.
In addition, Lough Ree was lowered in early summer this year by 100mm and this didn't impact negatively on navigation during the season.
However, it was a close-run thing; some heavy rain in July gave a boost to levels that were bordering on being too low for navigation in the vicinity of Meelick. This experiment will be maintained for a further two years, until the CFRAM data is available.
Despite farmer concerns, it does appear that all relevant bodies discharged their responsibilities during last year's summer floods. Still, I have an image in my head that won't go away – the derelict farmhouse on the riverbank at Clonbunny, and the family that lived happily there for more than a century.
Was this house simply 'the last place to flood'?