The International TOVS Working Group and the international atmospheric remote sensing community lost a friend and colleague with the passing of Paul van Delst on April 17, 2016, in Washington DC. Paul was known to many of us as a careful and conscientious scientist, and as a genuinely funny and free spirited individual who got along with everyone.

He first became known to this group in 1991 through the mentorship of Merv Lynch, his professor from Curtin University in Perth, Western Australia. Paul graduated with a Bachelor's degree with honors in physics from Curtin in 1989, and soon after got a job in the Curtin Physics Department doing materials testing. After slicing his finger with a scalpel while cutting a sample from a conveyor belt, Merv convinced Paul to try remote sensing instead and got him started on a PhD. As he was to do many times over the years with a number of his students, Merv sent Paul to the Space Science and Engineering Center at the University of Wisconsin-Madison, starting with a head first dive into the ITWG community at the 6th International TOVS Study Conference in Airlie, Virginia. During the conference Paul presented his PhD research on retrievals of ozone profiles using high spectral resolution infrared data; hit a home run in the USA vs. the World softball game; was made an honorary deputy by the local sheriff; and made a lot of new friends. Once he arrived at SSEC he was quickly recruited by Hank Revercomb to work on simulations of the double difference concept for comparing airborne high spectral resolution data to AIRS data. At one point Paul was proud to own the first Pentium PC in the center, and would gleefully explain the incredible performance of the 66 MHz CPU, and whisper that he couldn't believe that SSEC had bought it for him, as it had cost upwards of 10K. Paul honed his Fortran programming skills by developing his own code for radiative transfer simulations and retrievals, and by happily rewriting Bob Knuteson's code when Bob offered it for him to use for his PhD work. Paul Menzel decided to sell Paul his old motorcycle, and after patiently teaching him how to operate the machine, was horrified to see Paul ride off down the street on the wrong side of the road. Around this time Paul carried out a very careful comparison of GOES-8 sounder data against HIS data obtained from a NASA ER-2 over the Gulf of Mexico. He ventured into the field with the HIS crew on several occasions, and said he became a vegetarian after seeing all the doomed cattle in a stockyard in Coffeyville, Kansas, during an HIS deployment.

By the 2000s Paul had obtained his PhD from Curtin University, and was a fixture at SSEC, known to just about everyone in the building. He was right up there with Hal Woolf and Fred Nagle when it came to understanding the intricacies of Fortran programming (of course he was, they had schooled him). Paul also wrote a lot of analysis and visualization code in Interactive Data Language (IDL), and he was the one who convinced me I could write a book on the subject. We worked together on an interactive viewer for MODIS Airborne Simulator image data, and it showed us that the visualization of a dataset could be almost as important as the data itself. We would get coffee most mornings at the old Union South across the street from SSEC and discuss whatever tricky IDL or Fortran programming problem had presented itself the previous day. By this time Paul had made himself an expert in radiative transfer modelling via FASCODE, LBLRTM, and OPTRAN, and people in the data assimilation world started to notice this Australian guy who kept showing up at ITSC and seemed to know a lot about the subject. In the mid 2000s he was made an offer by NOAA/NWS/NCEP to come out to Washington DC to work on a new RT model known as the Community Radiative Transfer Model (CRTM). In time he became the unofficial spiritual leader of the CRTM development team, and shepherded it into the operational NWP forecasting system. He also worked very hard to make sure CRTM was open and accessible to others in the RT modeling community, a fact which led to his appointment as a co-chair of the ITWG radiative transfer model working group. Paul was an evangelist for high quality software development, and gave training courses at NCEP and elsewhere on best practices for collaborative software development. He was an extremely thorough (picky) programmer, and would agonize over a difference in the sixth decimal place in the output from the same program compiled by two different Fortran compilers. While lesser programmers might wrongly blame the Fortran compiler if their program behaved erratically, with Paul you could bet that if he said there was a bug in the compiler, then there really was a bug in the compiler. While living in DC, Paul had an old Toyota sedan from his days in Madison, but after he crashed it (his own fault, he admitted), he decided he didn't need a car, and from then on relied solely on bike, bus, or his own legs to get around DC. He would zip through some sketchy neighborhoods on his way from southeast DC to the office, and his main complaint would be the drivers who seemed to take no account of cyclists. On weekends he would get together with a group of friends and cycle a "century" (a hundred miles) in a day. He took several cycling tours through Europe and developed a taste for fine wine and food while traveling many tens of kilometers per day through France, Spain, and Italy. If he was visiting ECMWF or the Met Office, he would round up a bike and take off for a ride, leaving anyone who cared to join him far behind. Fortunately, he was the kind of guy who would stop and wait for you.

If you ever spoke to Paul, you would have noticed that he did not temper his Australian accent or slang for anyone. Phrases such as "same dog, different leg" would pop out in any conversation at the pub, at the office, or during a conference talk, and it made no difference to him. He was Australian and proud of it. Nevertheless, he did obtain a green card and US citizenship since he saw his long term career in the USA. He did have a little speed bump along the way when he received, after many delays, his permanent resident "green card". As he read it over, the birth date was correct, the name was correct, and the country of birth said "Austria". I can imagine the stream of colorful language that came next, as he resigned himself to more months of interactions with the immigration service to get it right.

In the last year or so, Paul expressed an interest in "maybe coming back to SSEC", if the conditions were right. Dave Tobin and I patiently encouraged him, and we were finally able to get him to agree to accept an offer to return to SSEC to work in a scientist position. He would have arrived in May 2016, and we already had tasks lined up for him to work on. More than that, we were just really excited at the prospect of having Paul around the place again. It just doesn't seem right that we won't get to see him again. Our community has lost a very good scientist, but more importantly we've lost a very good mate.

Liam Gumley
SSEC, UW-Madison

20 April 2016