David Fisher reports on how close New Zealand and coalition forces got to a daring rescue mission for Kiwi nurse Louisa Akavi.
Her codename was Raupo. Those who fought to free Red Cross nurse Louisa Akavi from Islamic State rarely used her name.
The Herald can now report details from inside the hunt for Akavi, which has seen confidence over her location and survival diminish over the week since it was learned she would lose the shield of secrecy presumed to have protected her since October 2013.
It shows the degree of focus and intensity which has consumed our diplomatic, intelligence and military over the years of Akavi's captivity. Among the team who sought to rescue Raupo are those who see her identification as a setback which could be fatal.
Raupo is a distinctly New Zealand word for a distinctly New Zealand operation. It is the name given to a wetland plant found growing at the edges of lakes, swamps and slow-moving rivers, and the fictional rural town from the Murray Ball's Footrot Flats comic strip.
The security and secrecy around Akavi's identity went to the heart of the intelligence assessments around her survival. Her captors made it clear from first contact that any publicity around her capture would be fatal.
• How New Zealand nurse Louisa Akavi was secretly held hostage by ISIS for almost six years
• Revealed: Keeping the secret of Isis hostage and nurse Louisa Akavi
• 'She's not once been forgotten': Friend of NZ nurse Louisa Akavi held hostage by ISIS pays tribute to 'amazing human being'
• Red Cross believes kidnapped Kiwi nurse Louisa Akavi is alive and will be found
In truth, Akavi's grip on life was tenuous from the moment she fell into the hands of Islamic State. Yet they would seize any advantage, however small, to aid her survival.
Over the last week, there has been concern among those linked to Operation Rocks - the New Zealand name for the mission to bring her home - that her survival has become even more unlikely.
This assessment follows the decision by the International Committee of the Red Cross to name Akavi and two others who were captured in October 2013.
The 'caliphate' is no more. But the Islamic State isn't finished yet.
US-backed Syrian forces declare victory over Islamic State
Shamima Begum was 'cruel Isis morality enforcer', aided suicide bombers
The Herald understands the Operation Rocks' intelligence assessments prior to her being named offered a handful of options. The grimmest was that Akavi had been killed in the assault on Baghouz, the Syrian town on the banks of the Euphrates which borders Iraq. It was the final geographical stronghold for Islamic State and it fell after weeks of intense bombing and fighting.
It was also considered she might have escaped Baghouz and was in refugee camps set up for the thousands of people who had fled Baghouz. This was considered possible although least likely.
Also possible was a "battlefield discovery", as the Operation Rocks team described it, which reflected finding Akavi alive in the ruins of the battle-scarred town.
The fear for her safety comes from the final option, which is that Akavi remains in captivity. The leader of Islamic State, Abu Bakr al-Baghdadi, personally held Akavi captive for years. He was not believed to have been in Baghouz when it fell, and it is considered possible Akavi - who was prized for her medical skills - was swept away from the final battle by his small group of senior followers.
A source linked to Operation Rocks says the risk to Akavi would have increased when this group - already the focus of a fierce manhunt - became aware of the intensity of the search for the Red Cross nurse.
The degree of effort is reflected in information provided to the Herald showing how our intelligence and military were inserted at key United States military posts which co-ordinated and drove the fight against Islamic State.
The Herald understands four NZ Defence Force personnel and an intelligence officer were specifically tasked to the region in 2016 to work on obtaining intelligence about Akavi's location.
Those in Operation Rocks became assured the presence of NZDF and NZSIS personnel in the region helped raise Akavi's profile and improve the chances of finding her alive.
The mission to recover her revealed how integrated our military has become with US-led coalition efforts. One key point in obtaining information leading to Akavi was Operation Gallant Phoenix, an intelligence-sharing taskforce set up in 2015 by the United States' Joint Special Operations Command and based at Amman, Jordan. Gallant Phoenix collected information gained through commando and special forces raids, from documents to DNA, with a wider brief of sharing information with partner nations so as to root out ISIL and its friends in their own countries.
By the end of 2016, it had grown to include more than 20 countries with 250 people, including two NZDF staff with a specific brief to obtain intelligence relating to Akavi.
There were two other NZDF staff based at the Kurdistan capital of Erbil, three hours from the Syrian border and just over an hour's drive from ISIL-held Mosul. The US set up a Joint Coalition Co-ordination Centre at Erbil in April 2015 to sharpen focus on the response to ISIL and the United States' Special Operations Joint Task Force was believed to have operated out of the city's airport. There was, in addition, a New Zealand intelligence agent tasked to work across the region.
By then, ISIL was collapsing at the edges. In December 2016, there was solid intelligence of Akavi's survival and a specific location which changed everything.
It was the first intelligence in a year to confirm Akavi was alive. It was the first time in two years New Zealand knew where she was.
Solid, trusted information kept coming. Akavi had been in Al Tabqah, on the south-eastern shores of Lake Assad. In mid-January 2017, she was moved to Raqqa, about an hour distant.
It is likely her presence in Raqqa, considered for a time ISIL's capital, provided the improved stream of information. It was a large city with a polarised population - many were suffering as a result of ISIL's holy war and Western intelligence agencies found fertile ground for sources.
Akavi's captors settled her into a safe house. The Herald understands she was moved twice in the following month - to a new safehouse and then back to the original.
The fresh intelligence spurred increased deployment and effort. Instead of four NZDF personnel, there were now eight people, and an additional three MFAT staff and an NZSIS officer.
By early 2017, the group was working from Erbil, which had developed as a base for US special air operations and was a fast-growing intelligence collection point.
At a ministerial meeting on March 13, 2017, the range of options available to recover Akavi appeared to have broadened as ISIL was pushed back.
The specific purpose of the meeting was to consider advice and intelligence from Five Eyes partners around options for rescuing Akavi.
The ministers at the meeting - Defence Minister Gerry Brownlee, intelligence agencies' minister Chris Finlayson and Prime Minister Sir Bill English - were told they believed Akavi was still alive. They were also told of the belief Islamic State was using her medical skills and moving her from place to place to meet their medical needs.
The briefing included effectively ruling out a planned military extraction by special forces. Raqqa was held by Islamic State and it would be months before the battle to defeat it began. At that stage, the city was well-defended and an assault by special forces simply wasn't feasible. It could be possible if she were moved out of the city, although it would require a fresh assessment depending on where she was taken.
If there was an urgent threat to Akavi's life, an "in extremis" operation was possible although unlikely. If it was to happen, it would be through coalition special forces, rather than the NZ Special Air Service troopers in the region. Even with imminent threat to Akavi's life, such an emergency operation was highly unlikely - it was simply too risky.
English, Brownlee, McCully and Finlayson were told the most likely option put forward was contracting agents inside the country, with ministers assured coalition nations had approved similar missions using a variety of groups acting locally. These included trained armed forces, smugglers and spies. Those forces would carry out reconnaissance, then attempt to quietly remove Akavi to a place where she would be met by coalition troops.
Such an operation needed to anticipate Akavi willingly accompanying rescuers, and being forcibly removed. Her state of mind, and the situation in which she existed, was unknown.
Intelligence officials and military special operations commanders, along with representatives from the diplomatic corp, were available at the meeting to answer questions.
It was impressed on ministers and the Prime Minister that planning such a mission could, in effect, lead to having to immediately carrying it out. The effort to confirm Akavi was present at the place where she was believed to be could expose the rescue attempt and require the plan being put immediately into action.
There was no half-measure option available - all in or all out. And if New Zealand was in, then there needed to be detailed study of the area in which Akavi was held, strong intelligence about who controlled it, the freedom of movement and level of security, and a route out to safety.
At that point, Akavi would be met by NZDF personnel - likely NZSAS troopers who had worked on the hunt for the lost nurse.
If it was to proceed, the ministers should understand such attempts were rare and extremely risky.
In the end, the risk was considered too high and Operation Rocks went back to tracking Akavi, looking for a fresh opportunity.
Incoming Prime Minister Jacinda Ardern was first briefed on November 3, 2017 but the Herald understands it was a February 2018 briefing from MFAT's Bernadette Cavanagh which raised the possibilities offered by the continued collapse of Islamic State.
The fluid, unpredictable field of conflict had allowed the team working in the region - and often inside Syria - to develop increasingly improved intelligence on Akavi's whereabouts.
Cavanagh's briefing pulled together information from MFAT, NZDF, the Ministry of Defence, the intelligence agencies and the Department of Prime Minister and Cabinet, the unifying cog in the Government's wheel.
The briefing was for Ardern, and also Cabinet ministers Andrew Little (intelligence agencies), Ron Mark (defence) and Winston Peters (foreign affairs). Peters was the lead minister for Operation Rocks - taking over from McCully - and MFAT the lead agency.
They were told there would continue to be greater opportunities to rescue Akavi, who was facing an increasing level of risk. To take into account the possibility for the sprawling conflict to change rapidly, the Inter-Agency Planning Team wanted greater scope to make and act on decisions at short notice.
It had been possible over the previous year to plot Akavi's movements and plan rescue operations. The presence of New Zealand personnel in the coalition hub also meant Akavi's survival was in the minds of those planning military action against Islamic State.
The New Zealand team believed they had tracked Akavi from Raqqa along the Euphrates River Valley, towards the Iraq border and inside ISIL's ever-shrinking sphere of control.
Now was the time, Ardern and ministers were told, to allow the team in Iraq and Syria greater flexibility and freedom to swing in behind operations which could result in Akavi's freedom, and to do so with a lower burden of proof as to her presence than previously required.
Such increased freedom would still not see the NZSAS used to carry out a rescue. If military force was required, then New Zealand's contribution would be to provide intelligence and planning support which would then see coalition special forces go into action. Four NZDF personnel - the NZSAS field teams in groups of four - would be ready to meet Akavi and escort her to safety.
Such a mission would likely take the NZSAS into areas beyond the control of coalition forces. It would see the team operating in extremely dangerous territory.
The speed with which events could unfold appeared to be the spur for the recommendation the Chief of Defence Force - then Lieutenant General Tim Keating - to be delegated authority to approve any mission which emerged.
Until now, Keating would have needed to seek Beehive approval.
Ministers agreed the Chief of Defence Force could now, if necessary, make the decision himself.
Efforts to recover Akavi also required specific legal consideration of any payment made to a third-party. A sum of US$200,000 was available to the Chief of Defence Force (now Air Vice-Marshal Kevin Short) to secure assistance if required, with NZDF and MFAT lawyers making it clear there could be no payment made to anyone who was a member of a terrorist organisation.
Part of the consideration was to avoid any public perception there had been a ransom payment made.
The changes made showed Operation Rocks had adopted an increasing forward-leaning stance. It was as if it were on starting blocks for a sprint to the finish - the initial settling into race position had been raised from "ready" to "steady" and all involved were braced for the starter's pistol.
The Coalition fighting ISIL had all but destroyed the physical Islamic State, reducing its territory from spanning two nations and 10 million people to the town of Baghouz.
All the preparation work appeared to be on the verge of bearing fruit when Akavi was one of three hostages offered as a bargaining chip for high-ranking members of Islamic State seeking to escape the town.
Nothing eventuated, and Akavi did not emerge - not from the town, when it fell, and not as a casualty of the final battle.
It was in this vacuum that her employers, the ICRC, went public with her name and the long-suppressed story of her captivity.
"It absolutely remains the Government's view that it would be preferable if this case were not in the public domain," said Prime Minister Jacinda Ardern.
The decision to do so was considered a setback for the Operation Rocks team. They considered her captors - if she remained in captivity - would now be rebalancing the value they saw in Akavi against the risk of keeping captive someone so keenly sought after.