Interview with Colin Simpson (CS) was head of Remote Sensing Research Group and Chief of Division of Environmental Geoscience and Groundwater at the Bureau of Mineral Resources (BMR)/Australian Geological Survey Organisation (AGSO)

by Catherine Rayner (CR)

on 11 October 2001


(This is not a verbatim transcript as many asides, comments and remarks have been left out but essentially the majority is “as spoken” – Paul Wise)



CR: What I'm interested in is you and your background, how you became interested in remote sensing, why you stayed with it, why you enjoyed it, those sorts of things?


CS: There's not much doubt that my interest started when I was at Melbourne University doing geology and one of the courses available was geomorphology presented by (unrecognisable). The surface geomorphology class had in it a very large component of air photo-interpretation. One of the few geology courses in Australia at that time that did have a lot of air photo-interpretation in it. I really enjoyed working with aerial photographs. After I left university I joined Mt Isa Mines in exploration and I stay with them for five years. I then heard that the Bureau of Mineral Resources (BMR) was looking for a person with photo-interpretation experience so applied for the job and got it.


BMR had what they called a Photo-Geology Group and that was the main support group for their mapping programs. BMR’s primary role at that time was to complete the geological mapping of Australia at 1:250,000 scale. The most efficient and effective method that they had for mapping was to employ aerial photography, interpret aerial photography, them home in on key areas of interest or importance they identified on the air photographs. To help speed up that process they set up the photo-geology group of three people whose sole task was to interpret the aerial photographs of the whole area that was to be mapped in the subsequent years. They would interpret the geology across all of the photographs that would be then compiled into a map, a very preliminary photo-geological map, and the geological team that went into the field to do that mapping would already have a good understanding of what to expect.


So the prime task of that group was to simply interpret the photographs, get as much geological information onto the face of the maps as they could and then they would move on to the next set of maps for the subsequent year.


I joined that group in 1968 and in a matter of only three or four years after that there was a whole range of what came to be known as remote sensing technology became available. Because we had been involved with the interpretation of aerial photographs the task fell to that group to undertake and assess the potential application of new ranges of remote sensing instruments that were becoming available, to see how they might be used in our geological mapping program or in BMR in general.


So that's how I came in to it. The technology that became available comprised both declassified military equipment, such as side-looking radar and airborne thermal scanners, that were employed in commercial fields for mapping features on the ground, and then of course the first Earth Resources Technology Satellite (ERTS) was launched in the early 70s. One thing led straight into another and from that point on there was an almost continuous proliferation of new technology coming through that had to be assessed, evaluated and trialled. If it was appropriate we applied it in our programs. So I got into it in that fashion and because it was very interesting and exciting time, in terms of science and technology, and because there were new things coming along all the time, it was a great world to work in. I simply stayed in it and got up to a point where I was Principal Research Scientist in remote sensing in BMR and the subsequently formed Australian Geological Survey Organisation (AGSO).


I should make the comment, as I think it’s quite relevant, that from my point of view in geological mapping, Australia had the most efficient geological mapping program in the world at that time in the 50s, 60s and early 70s simply because it did employ aerial photography. There were not that many countries in the world at that time that applied aerial photography to the full extent that Australia did. America lagged dramatically behind Australia and even some other countries in their use of aerial photography for geological mapping. Aerial photography was really the key that allowed Australia to map such large areas of this country so rapidly with a relatively small team of people.


The photography program started up primarily after the war when the technology became available. There was quite a flurry of activity during the war taking aerial photography in strategic areas and that was co-ordinated both through Defence and also at that time through National Mapping. National Mapping also used aerial photography to the fullest in the preparation of all their primary topographic maps of Australia. So it was a fairly easy step after the war, when the technology became available, for them to put in major programs that collected aerial photography all over Australia and as far as they could they also collected aerial photography in Papua New Guinea (PNG).


When the first new remote sensing technology became available one of the first things that was picked up and applied in PNG in particular was side-looking airborne radar (SLAR). The reason that it was applied in Papua New Guinea was that the very extensive cloud cover particularly around the mountainous areas prohibited good aerial photography. So it was difficult to produce the maps that were required. But the side-looking radar was able to penetrate the cloud layer and was an ideal tool and there was complete coverage acquired by National Mapping over the whole of Papua New Guinea. There were a few lines of it flown in Australia mainly for experimental or research purposes.


When the first ERTS satellite imagery became available it was a fairly easy step for Australian researchers and scientific organisations, involved in mapping for a variety of purposes, to recognise the potential value of these photos that covered large areas of country. We weren’t terribly excited about what we saw on the photos in respect of the distribution of geology or distribution of vegetation etc because that was typical of what we had utilised in our aerial photography. What we did find particularly exciting in the first instance using those [ERTS] photographs was the ability to map traces of major fault lines and major tectonic features, which were simply known as lineaments. That was the terminology which tended to come up and be applied very strongly in the early days of the ERTS satellite imagery because that large area coverage did allow you to see features that would go all the way across a [ERTS] photograph. By lining up bends in rivers, erosion scarps, changes in soil type, which were obviously structurally controlled features related to sub-surface geology, we were able to map out enormous amounts of tectonic information on these images that we had not been aware of before.


Now when the ERTS imagery first became available in America it virtually revolutionised their approach to geological mapping as it was the first time that they had seen large coverage photographs of the earth whereas we had been using it for many years. Anyway with the enthusiasm that came out of America for this imagery it made sure that that program rolled on resulting in a number of follow-up satellites over the years.


I guess the next big step that we saw was the change from straight photographs as produced by the first satellite, to using that information in a digital form. I think there was a period of about three years after the launch of ERTS 1 where we didn’t see any digital data at all. In fact I don’t think the Americans (NASA) released digital data early in the piece as there were very few people capable of utilising it. When digital data did become available, within about two years, which would have been around 1975/76, I guess only 5 agencies, at the most, in Australia would have been capable of utilising it. There would not have been any more because you needed not only very extensive computing capability, but you needed people that could do it. The sort of computer and image processing software and hardware, which was available, was something in the order of $250,000 to acquire. You had nothing at all like the current day PC availability back then, so all the computer processing that went on was in a mainframe environment and there were not that many people ready to jump straight into that sort of environment; CSIRO did, BMR did, National Mapping did, and exploration companies like CRA and Western Mining got into it fairly early, but there were very few.


Therefore, after the launch and availability of the first imagery the next big step was really the advent of digital technology and where it’s led to now.


What I’ve spoken about largely so far is simply the ERTS, but there was a whole series of other space activities that went on such as Skylab. Skylab produced some very nice imagery, equivalent to high-resolution aerial photography, across Australia at various sites. Australia was a Principal Investigator for NASA in assessing Skylab photography, and there were a series of other satellites with various capabilities which came available about that time.


CR: Perhaps we could wind-up this section with what you’re doing now?


CS: I left AGSO in about 1996 simply because of the changes in the public service; it was “inefficient” for me to stay. So I left and set up my own consulting company and over the last five years have been involved in consulting, some of which has been in remote sensing, specifically remote sensing advice. Not in hands-on processing per se but recommendations and advice. Most of the activity has been in environmental geoscience as in the last five years of my time with AGSO I was Chief of the Division of Environmental Geoscience in Groundwater. So I had actually moved out of very active remote sensing investigation, research and application and into a senior management position. That was the first environmental geoscience activity formally set up. Since I left I also have continued my attendance at international conferences on geoscience.


Australia viewed the first satellite imagery very differently from that of America. The Americans were very enthusiastic and over the moon and when they (NASA) came to Australia to the very early conferences, as I recall, they didn't see the great exuberance in Australia about this imagery to the same extent as they had seen in the US. They were a little bit puzzled and amazed at that, they expected everybody to be jumping up and down. From our perspective the photographic information in there wasn’t nearly as good as what we had got out of aerial photography that we'd been using for many years. The structure of tectonic information, however, that was in the imagery was really great and that was what a lot of the early research in Australia focused on. Until we got digital information and digital capability and were able to discriminate digitally the basic photographic content, to us, wasn't terribly exciting.


CR: It sounds like in fact Australia was ready for much more sophisticated use of the data.


CS: Yes very much so. In fact that’s one of the reasons that when we did get hold of digital data, and it only took three or four years (the CSIRO group under Ken McCracken had specifically enlisted the involvement of Jon Huntington and Andy Green) to become the fore-runners in the world for processing digital date for geological information. No doubt they were far ahead of anyone else, because they were not just interested in just looking at it and wanting to create some sort of high discrimination between vegetation, soils and other things you could do. They were interested in getting in to process the data and trying to pull out a lot much more subtle information than the Americans were going for because we could get a lot of the other information from ordinary photo interpretation.


They developed algorithms and processing routines that were far ahead of their time and a lot of them are still used as basic routines. You are probably aware that both Jon Huntington and Andy Green were awarded the Australia Prize and a lot of it came as a result of that very good research they did in the formative years on the digital side of things. They were part of Ken McCracken's team and the enthusiasm and potential that Ken saw in this sort of technology was such that he threw incredible support from his Division into it. He had people with the vision like Andy and Jon who could step straight into the required roles of image processing and interpretation. Just getting a basic understanding of just what the data was and how it operated so that they ended up making some very major advances.


I just want to draw your attention to Driscoll’s, Western Mining Corporation, work in remote sensing. He put together plots of major lineaments, linear features on the earth’s surface, from magnetics, gravity and topography and also from what he could map from aerial photography and satellite imagery. He combined them all of these trying to establish working relationships between these features which indicated structures in the earth and mineralisation. In South Australia, he ended up with linear patterns one of which was near a known mining field. He came up with a subsequent set to the north. He targeted this northern site as it looked almost identical and had the same characteristics but there was no known mineralisation there, but took out an exploration licence. Drilling where he had predicted the very first drill hole hit mineralisation at about 600m or so. There was no outcropping or other surface indication whatsoever.


I don't know how much of the lineament evidence came from satellite data rather than magnetic or gravity data, but satellite data was definitely in it and Roxby Downs is one of his pet areas. Western Mining still play their cards very close to their chest but still use this technology although he has been retired for several years but he, Kerry O'Sullivan and Lyle Burgess would always turn up at remote sensing conferences and always add significant input. The mining companies certainly had some “clout” and helped in getting the Australian Landsat Station built.


CR: I am interested in the interaction between government, the companies, semi-independent organisations like CSIRO, how they all pushed each other and what they fed each other in terms of information?


CS: There was a lot of very good into action between the companies and the ACERTS committee etc. 


CR: Can you think of any other big events or changes that really stick in your mind?


CS: One significant event was the establishment of the Australian Committee on the Earth Resources Technology Satellite (ACERTS) committee. The Australian government departments such as the Bureau of Mineral Resources and the Division of National Mapping were very much in favour of Australian being involved in acquiring, particularly spacecraft related imagery, because of its relationship with aerial photography and the setting up of the ACERTS committee was a very specific factor in that.


Certainly it didn't take too long for the ACERTS committee and others involved to see that Australia needed its own receiving station. Simply because if you didn't have a receiving station you could only acquire data, such as over Australia, by having certain areas imaged onto the on-board tape recorders and they had a very limited capacity. This meant you could only get a few images per day of any site because that data had to be recorded and then when the spacecraft was over the United States or Canada it was downloaded to receiving stations there and the tape was automatically wiped so it was available for the next lot. That meant that we had only a very limited capability to acquire images over various sites in Australia. It became apparent very early on that the problem with the life of the tape recorder wasn't guaranteed for very long and we had this great big continent over which we wanted imagery and we were never going to get it by relying on the on-board tape recorder. The only way we could do it was to have our own receiving station, centrally located in Australia at Alice Springs, which was an ideal site because it would give us complete coverage of Australia and up to most of the relevant areas of Papua New Guinea. That became a major push so that the acquisition of a receiving station I would see as a major breakthrough.


The other major breakthrough was certainly, from a remote sensing satellite perspective, was the availability of good digital data for digital manipulation. Certainly the achievements of CSIRO in data manipulation and algorithm development, as I mentioned before, was particularly significant. I guess from then on, the other real major achievements have been things that have spun out of all that. For example, it was apparent very early on that we would have to get into other wave-length bands and get satellites up there with a lot of sensing bands in very specific areas that were relevant to the Earth's surface. The aircraft research and aircraft systems that were developed that actually filled in those roles became the next major significant activity and of course we are now starting to see those systems. The sort of systems we envisaged probably 10 or 15 years ago are now starting to come along, and while it's been a long hard road, then again it is a very expensive exercise to build hyperspectral systems to go on satellites. The hyperspectral systems which have 250 bands or thereabouts are aircraft operated and proving exceptional value in a whole range of activities from minerals exploration through to environmental management. The hope is that this type of progress keeps going as there is still the cloud out there for people to convince government agencies such as NASA to keep looking and finding (unrecognisable).


The growth of private satellite systems has been beneficial but has had a downside not the least of which is that some systems have collapsed or disappeared and the capability to receive some of the new systems has presented a bit of a problem as you have to get funds to continually upgrade and modernise your receiving station. That means that the whole science of convincing governments of the need must continue on all the time. You don’t need much of a lapse or diversion on the government’s part in these things, and you can be put back for some considerable time.


CR: I was speaking with Ken (McCracken) about the X-band upgrade to the ALS which was a nice example of that. If they didn’t get that, they didn’t get data, so what do you do?


CS: Yes you’re right. I think there is another area of the whole scene of remote sensing in Australia which has been a very significant problem, and that has always been the problem, I guess summing it up in one word,  “education”. The difficulty of information transfer from those people who are aware and know of the capabilities of remote sensing into those areas, non-technical or sub-technical areas, in government agencies and companies where the use of that information and technology can be extremely cost-effective. The big difficulty has always been, and continues to be, the process of educating people.


Now you can have Universities doing it whereby they educate people coming through the system and the UNSW has been a very significant player. Probably the University at the forefront of promoting remote sensing. John Richards played a major role both at ADFA and before in the education role. But there is still the hiatus in there. It is one thing to educate the student coming through the system about it, but the people who have to be educated a lot more are the existing decision makers in government and in all actual areas of applications. That has been something which has dogged us ever since remote sensing effectively came into being.


It is still a very significant problem. Until you can get the knowledge available into government agencies and departments that here is a tool that is extremely cost-effective and can help you to make management decisions in whole range of scientific disciplines and areas; management decisions which have implications for the long term and for monitoring; until you can convince those people then you can’t build up a strong political base that you can utilise to get the government to continue to make adjustments to Landsat stations or to do upgrades or take on a new technology that will be the next step and improve it.


It takes quite a considerable time between when the technology is actually developed and when the applications are fully known. Generally as it takes a matter of some years for it to come in, you have difficulty in maintaining a lot of the impetus in the political background to it. Education is probably one of the biggest difficulties the whole remote sensing field has faced. How to get across particularly to the non-scientific decision makers that it is a tool that is incredibly cost-effective. The trouble is when they look at satellite information now and a program to process it on their PC they see $1000 per picture, they say “We’re not going to pay that”!  But the information that you can get out of that picture far surpasses anything else that you can get on the ground by any other means. If you can then feed that information straight into long term management or environment applications or whatever you want to do with it, it’s a cost-effective business. Convincing the non-technical people of the cost-effectiveness of it has been the biggest stumbling block.


CR: Has there been any co-ordinated effort by the industry to do so?


CS: The Remote Sensing Association in Australia has tried to get co-ordination but then they’re relatively small players in the whole game. They have a lot of very interested and enthusiastic people, but are still relatively small in numbers. At the end of most remote sensing conferences when summing up how successful that conference might have been or what needs to be done you almost invariably find the same comment coming up that we need more education into the users domain. While the researchers in Australia have been extremely good and world leaders, the actual application transfer of that knowledge into the non-scientific areas or outside of the domain of research has been fairly poor and I honestly don’t know how you really push that vigorously.


The Landsat station runs education programs of that sort and has its newsletters. There are programs that are run through the Australian Mineral Foundation, primarily aimed at educating the mineral industry sector on the applications and AMIRA has programs as well which are industry oriented. I honestly don’t know, unless you have a whole lot of money to spend which is something that the whole system doesn’t seem to have, how to put in a program that can reach out to the non-scientific areas.


Even in areas in the Department of Environment for example that are not necessarily being run by scientists, if the term remote sensing comes through the literature it won’t mean anything to them. Even if it does they still have to be able to bridge the gap between what they see in glossy brochures and what its relevance may be to their sections. It really is an incredibly difficult area.


CR: That’s why there is a Centre for Public Awareness and Science at ANU that looks at this sort of thing.


CS: It is a problem relevant to all modern technology and science and I am not sure how you overcome it.