I work on infrared astronomy and cosmology.
Some details of my current and recent research are given below. In 2014 I gave a 6-lecture postgraduate course on ‘Dusty galaxies’ at the University of Bologna (text of lectures here).
I am also strongly involved in astronomy outreach activities, both in my home locality and more widely. I live in Southwold, Suffolk, and write a monthly column in the Southwold Organ about astronomy, tides, coastal erosion and climate change called Stars'n Tides. I am Honorary President of the Darsham And Surrounding Hamlets (DASH) astronomical society. I have given talks about astronomy and space to primary schools and worked with local cubs for their astronomy badge. I gave lectures at the Thessaloniki Demetriad Festival in 2008 and 2009 (texts here). I took part in a debate about the reality of black holes at the Hay Science Festival in 2015 (my contribution here). I gave a talk via webinar in Nov 2015 to students at Kharagour, India, on Astronomy from Space.
I have also become increasingly interested in the history and philosophy of science. I gave a talk in 2013 at University College London on 'Reflections on Kant and Herschel: the interaction of theory and observation' (text here, published in Harmony of the Sphere, ed. Silvia De Bianchi, 2013, Cambridge Scholars Publishing).
I gave a talk on 'Shakespeare's astronomy' on the BBC World at One on Shakespeare's 450th birthday and a longer version as part of the IXth INSAP (Inspiration of Astronomical Phenomena) conference in 2015 at Gresham College, London (text here.)
I have a strong interest in Aristotle and his legacy. I wrote an article 'Was Aristotle the First Physicist ?' in Physics World in 2002 (text here). Work in progress is to expand this to book length: 'At the Lyceum of Aristotle'.
I gave a lecture on 'Interplay between Evidence and Theory in Astrophysics and Cosmology' at the Institute of Advanced Study, Durham, in 2015. Examples discussed included Copernicus, Galileo, Maxwell, Einstein, Hubble (text here).
My 'Contemporary Authors' entry is given here. As well as writing on astronomy and science, I wrote a play 'The LIfe and Times of Nicolas Koppernick of Torun', which was broadcast in a Polish translation by Bolek Taborski by the BBC Polish Service in four parts during December 1973 as part of the celebrations of the 500th anniversary of Copernicus's birth.
Lingyu Wang and I generated the Imperial IRAS Faint Source Catalog redshift catalogue (IIFSCz) in 2009. We have now produced a revised version of this (RIFSCz), incorporating SDSS, 2MASS, WISE and Planck data (Wang et al, 2014, MNRAS 442, 2739). A further update (Nov 2014) incorporates Akari data and improved optical and near infrared data for 1271 nearby galaxies, see readme.
The SWIRE consortium (PI Carol Lonsdale, Deputy PI Michael Rowan-Robinson) consisted of astronomers from IPAC, UCSD, Imperial, Sussex, IAC Tenerife, Padova and other institutions in the US and Europe, and has carried out a Legacy Survey with SPITZER of 49 sq deg of sky.
A simple and versatile parameterized approach to the star formation history allows a quantitative investigation of the constraints from far infrared and submillimetre counts and background intensity measurements (Rowan-Robinson 2001).
Details of models for starbursts are described in Efstathiou, Rowan-Robinson and Siebenmorgen (2000, MN 313,734).
Our paper on cirrus models for local and high redshift cirrus models can be found here.
I first worked on models for the zodiacal dust cloud with my group at Queen Mary College, London, during the period 1988-93. Our goal was to model all the known components of the infrared emission mapped by the IRAS satellite during its 1983 survey of the 10-100 micron sky. Our final paper (Jones M.H.and Rowan-Robinson M., 1993, MNRAS 264, 237) included a model for the asteroidal dust bands at 1.5-3.1 au and a model for the extended fan distribution at r < 1.5 au, but not any contribution due to cometary dust.
During the summer of 2006 I became aware from press interviews that Brian May was interested in the idea of returning to the thesis on zodiacal dust that he had been pursuing at Imperial College in the early 1970s when his rock band Queen became a more pressing interest. I suggested to him that he come and talk to me about this if he was seriously interested. He sent me a write-up of what he had done in 1970-74 and we met up to talk about it. My view was that he had been very close to being able to submit his PhD in 1974. To submit in 2007 he would have to review everything that had been done on zodiacal dust in the interim and review how the anomalies that he had found in his 1970s work could be tested in the future. Within a year he had submitted his thesis and this was awarded in 2008.
The most interesting idea in the thesis was the suggestion that anomalies in the velocities of the zodiacal dust that Brian had measured in the1970s could be due to interstellar dust flowing through the inner solar system. We decided to see whether a paper could be developed from this. Initially we hoped that a new ground-based programme of measurements of the kinematics of interstellar dust could be developed, in collaboration with other groups. I also started to play with the 1993 models to see whether an interstellar dust component could be added. I knew from work by several groups that the strong contribution of cometary dust had to be incorporated, especially beyond the asteroid belt.
Our final model was published by Monthly Notices of the RAS in January 2013 (paper here).The fan extends to 1.5 au and is supplied by asteroidal dust and dust from Jupiter-family comets. These tend to have inclinations < 30 degrees, aphelia ranging from 5 to 30 au, and have their origin in the Kuiper belt. Interstellar dust is modelled as a uniform density, isotropic cloud of dust extending from the sun out to 30 au. 70% of the dust crossing 1.5 au is due to cometary dust, 22% due to asteroidal dust, and 8% due to interstellar dust. However only 1% of the zodiacal dust arriving at the earth would be of interstellar origin. We fit our model both to the IRAS data and to data from the DIRBE instrument on the COBE mission (1989-90). The latter is important because DIRBE had an absolute calibration and this is crucial in any search for an isotropic component. Sample fits to the IRAS (left and centre) and COBE (right) data are shown.
In 1960, while working at the National Physical Laboratory for a pre-university year, I became interested in the theory of numbers and studied the famous book by Hardy and Wright. I worked on partitions, a topic that had fascinated Ramanujan.
The partitions of an integer n are the number of ways it can be written as the sum of integers, without regard to the order. So the five partitions of 4 are: 4 = 3+1 = 2+2 =2+1+1 = 1+1+1+1, and p(4)=5. For n=1-10,
p(n) = 1, 2, 3, 5, 7, 11, 15, 22, 30, 42.
Partitions were extensively studied by Euler, Jacobi and Ramanujan. In 1918 McMahon published a table of partitions to n=200, with p(200) = 3972999029388. Gupta (1939) extended this table to n=600 and Gupta, Gwyther and Miller (1958) extended this to n=1000.
In my 1960 work I found a triangle for calculating partitions, analagous to Pascal's triangle for
binomial coefficients. I published this in Eureka 24 (1961, p.14). The triangle is based on the
p(n,r)-p(n-1,r) = p(r-1,2r-n-1)
where p(n,r) is the number of partitions of n with largest integer (n-r). A partial proof was given
in my Eureka article and this was completed by J.A.Tyrrell in Eureka 25 (1962, p.5). p(n) is then
the sum of p(n,r) from r=1-n.
Seeing the film The Man Who Knew Infinity inspired me to write a simple program to calculate
p(n) using this recurrence relation. Using the integer*16 option in gfortran I was able to calculate
p(n) to n=1437. The table of partitions is given here. The values to n=600 agree with Gupta's
table, which is available online. The values of p(1570,r) are tabulated here. The time taken to
calculate these tables was much less than 1 second. The method involves of order n2 additions.
The only limitation is on the size of integer that can be stored in gfortran (one could go further in
C). The value of p(1437) is the impressive number
p(1437) = 168434321304033467550147269349447360294
However Wilson(2006, oeis.org/A000041/b000041.txt) has tabulated p(n) for
n = 1-10000 and Wilf(2000) reports that p(n) has been computed for n into the
billions. I found 1000 mathematical papers with 'partition' in the title between
2002-2016 and there are still unsolved problems, for example the randomness in
whether p(n) is odd or even.