By Roger Harrabin
Environment analyst, BBC News
Phil Jones, the professor behind the "Climategate" affair, has admitted some of his decades-old weather data was not well enough organised.
He said this contributed to his refusal to share raw data with critics - a decision he says he regretted.
But Professor Jones said he had not cheated over the data, or unfairly influenced the scientific process.
He said he stood by the view that recent climate warming was most likely predominantly man-made.
But he agreed that two periods in recent times had experienced similar warming. And he agreed that the debate had not been settled over whether the Medieval Warm Period was warmer than the current period.
These statements are likely to be welcomed by people sceptical of man-made climate change who have felt insulted to be labelled by government ministers as flat-earthers and deniers.
Professor Jones agreed that scientists on both sides of the debate could suffer sometimes from a "bunker mentality".
He said "sceptics" who doubted his climate record should compile their own dataset from material publicly available in the US.
"The major datasets mostly agree," he said. "If some of our critics spent less time criticising us and prepared a dataset of their own, that would be much more constructive."
His colleagues said that keeping a paper trail was not one of Professor Jones' strong points. Professor Jones told BBC News: "There is some truth in that.
"We do have a trail of where the (weather) stations have come from but it's probably not as good as it should be," he admitted.
"That's similar with the American datasets. There were technical reasons for this, with changing data from different countries. There's a continual updating of the dataset. Keeping track of everything is difficult. Some countries will do lots of checking on their data then issue improved data so it can be very difficult. We have improved but we have to improve more."
Professor Jones clarified later that when he had told me that his paper trail was insufficient he meant data trail. He insisted that he had not lost any original data, but that the sources of some of the data may have been insufficiently clear.
His account is the most revealing so far about his decision to block repeated requests from people demanding to see raw data behind records showing an unprecedented warming in the late 20th Century.
Professor Jones said climate scientists needed to do more to communicate the reasons behind their conclusion that humans were driving recent climate change.
They also needed to be more transparent with data - although he said this process had already begun.
He strongly defended references in his emails to using a "trick" to "hide the decline" in temperatures.
These phrases had been deliberately taken out of context and "spun" by sceptics keen to derail the Copenhagen climate conference, he said.
And he denied any attempt to influence climate data: "I have no agenda," he said.
"I'm a scientist trying to measure temperature. If I registered that the climate has been cooling I'd say so. But it hasn't until recently - and then barely at all. The trend is a warming trend."
He said many people had been made sceptical about climate change by the snow in the northern hemisphere - but they didn't realise that the satellite record from the University of Alabama in Huntsville showed it had been the warmest January since records began in 1979.