Next week on monday and tuesday, there will be a spatial copula workshop, held by the fine folks at the Institute for Geoinformatics at the University of Münster. I’m looking forward to discuss copula-related issues with interesting people.
Here is some more information regarding my talk:
|date:||Monday, September 22nd|
|location:||Institute for Geoinformatics, University of Münster|
|title of talk:||“Spatial Interpolation Using Secondary Information and Censored Measurements ”|
|authors:||Claus Haslauer, Theresia Heißerer, András Bárdossy|
On Wednesday I will be giving a presentation in Paris, France. In the building where Georges Matheron worked. The conference “geostatistics for environmental applications” celebrates its 10th repetition.
If you happen to be in or near Paris, let me know — here are some details about my presentation:
|date:||Wednesday, July 9th|
|title of talk:||“Realistic Non-Stationary Spatial Interpolation”|
|session:||Geostatistical Theory & New Methodologies 1|
|authors:||Claus Haslauer, Theresia Heißerer, András Bárdossy|
We’ll be next door!
Photo by victortsu – http://flic.kr/p/ht1wM4
Here are some reflections on EGU 2014 that I started to write down on its final day, but haven’t had time to post until now.
There are only two things that stand out from the talks and the posters (Hazel Gibson collected a few extraordinary posters here).
- the best scientific talks are the ones that explain one thing well and then show a few results. I find that much more is not possible in 13 minutes. Particularly, I remember E? Foufoula, who had a beautiful slide that explained how she thinks about similarity and dis-similarity in time-series.
- in the posters I had great fun wondering around all kinds of topics. I found out how wind-turbine engineers try to place turbines optimally in a wind-turbine field. I learned about surface run-off. I had a great chat with Anna Sciani on infiltration processes at a hillslope, and about the worth and work related to the combination and cycling of and between field-experiments and associated modelling.
Besides the talks and the posters, I was particularly impressed by this year’s medal lectures that I attended.
Among the three, personally, I was moved and inspired greatly by Upmanu Lall’s talk. Bruno Merz said in his laudatio that he appreciates that Upmanu Lall can think outside the box. I remember that I thought during the laudatio that this seems like a weird thing to say in a laudatio. Now I think this was the best thing to say.
Here is the list of the three medal lectures I attended. Interestingly, Hoshin Gupta and Upmanu Lall are also co-author of recent discussion papers that appeared in Water Resources Research that are also linked. The papers contain at least some aspects of the talks.
- Hoshin Gupta on “Using Models and Data to Learn” (Water Resources Research paper)
- Eric Wood on “The Challenges of Developing a Framework for Global Water Cycle Monitoring and Prediction”
- Upmanu Lall on “Does hydrology have a soul?” (Water Resources Research paper)
at the beginning of the conference I was intrigued by the concept of a “scientific debate”. I attended the one on geo-engineering (also because there were no particularly seemingly interesting topics for sessions in the program — a rare occasion). The mindsets of the panelists were not very diverging, hence the debate was fairly calm. Here are my take-home messages:
- in an ideal world, if we had a crop that has the same nutritous properties, needs less water, is somehow generally better for the environment than existing crops, and leads to more yield, then yes, we would all be for that, and we could call that geo-(bio-)-engineering. And partly, this is happening.
- win-win situations are rare, and the big question seems to be how to “properly” “treat” the non-winners. Related to that, another big question is at what point in time and under what circumstances would we accept some large scale engineering project, with associated large, partly unknown, and uncertain consequences (“tipping point”)
- large scale vs. small scale: the discussion was largely focussed on larger-scale engineering works. At the end was some discussion if small scale or bottom-up approaches would not be better. But then, it was not clear how to know which small approach would have large consequences, and how it could be adopted to improve the consequences.
- the panelists, scientists, seemed more concerned about geo-politics than about geo-engieneering. Everybody agreed, that solutions in harmony between all parts of society, i.e. scientists, engineers, lawyers, economists, social scientists, medical experts need to be found.
At the google booth I was re-introduced to their GIS-related computing capabilities, called “earthengine“. It seems like anybody can cooperate with them for smart, computing intensive remote sensing type calculations and analysis. On their website they have some pretty impresive videos. This one here is about the growth of irrigation in Saudi Arabia, as evident from an explosion of pivot-irrigated plots.
Here is an interesting chart, that I saw yesterday at the IPCC session at EGU.
It shows the energy stored in different compartments. There is more heat stored in water than say in the air, due to the larger heat capacity of water. The exchange of heat, I think, must occur via temperature. So this must have an affect on the oceans, the living beings within the oceans, and on climate.
As Steve Easterbrook points out at azimuth:
The oceans act as a huge storage heater, and will continue to warm up the lower atmosphere (no matter what changes we make to the atmosphere in the future).
Description of this figure from IPCC
(Box 3.1 Fig 1) Plot of energy accumulation in zettajoules within distinct components of Earth’s climate system relative to 1971 and from 1971–2010 unless otherwise indicated. Ocean warming (heat content change) dominates, with the upper ocean (light blue, above 700 m) contributing more than the deep ocean (dark blue, below 700 m; including below 2000 m estimates starting from 1992). Ice melt (light grey; for glaciers and ice caps, Greenland and Antarctic ice sheet estimates starting from 1992, and Arctic sea ice estimate from 1979–2008); continental (land) warming (orange); and atmospheric warming (purple; estimate starting from 1979) make smaller contributions. Uncertainty in the ocean estimate also dominates the total uncertainty (dot-dashed lines about the error from all five components at 90% confidence intervals).
A good discussion about this topic, also related to the uncertainties in the predictions related to stored heat can be found at Climate Etc.
Everything is ready for EGU: the talk is prepared — see you all tomorrow morning!
Here are some details:
|date:||Monday, April 27th|
|title of talk:||“Using Secondary Information for Realistic Interpolation of Groundwater Quality Parameters and Associated Uncertainty Quantification”|
|session:||GI1.5 Applications of Data, Methods and Models in Geosciences|
|authors:||Claus Haslauer, Theresia Heißerer, András Bárdossy|
I remember from last year, that I enjoyed the spot pictured below for lunch: it’s outside, with a few on water (it might be a side-canal of the Danube, and: there is a physical map of Europe with typical rocks of different parts of Europe glued to their proper location. Hence, I am looking forward to a collection that the EGU has in mind, as they write here:
All participants are welcome to bring a stone from their home country to be collected in the entrance hall, placed in a showcase, and their origin will be marked on a world map. This will be a piece of geoscience art created by all participants together and is part of the 2014 theme Face of the Earth. Please catch up on the regulations in your country regarding the export of such a stone. The stone should not be larger than 12 x 12 x 12 cm. The artwork will be donated to an Austrian educational organization. Therefore, your stone will stay in Vienna. Exhibition Spot: Entrance Hall – Ground Floor (Yellow Level)
The mac has recently had its 30th birthday. I am too young to remember its first birthday, but I am old enough to feel the pain! I’ve never used anything pre OSX. One piece of software (as apps have previously been called) has been with me since the beginnings: omnioutliner. I won’t get into fights of outlining versus mindmapping. I think both have their place, and I also think that mindnode is an awesome software.
Taking notes that contain equations has been and is my bread and butter. When laptops were finally portable enough and affordable enough to be taken into a classroom, this was a joy for me, because I could use then omnioutliner for my notes — back then, a little app called “oo3eq” facilitated the combination of equations and omnioutliner. This is how it worked:
- type your equation in latex into the note field of omnioutliner
- hit a button, and a pdf would appear in the row that belongs to the note
- the equation would still be in the note, which I found very useful for future reference.
- having a pdf of the equation visible directly above the note, was much more pleasing to the eye, and helped following the notes that I took tremendously.
I did try to reproduce this behaviour, and you can check it out on github. There is an applescript that gets the content of the selected note as a string, which must be a valid latex math expression. This string is passed to a python script that generates a valid latex file, which is processed in the script. The result is a pdf file. Back in applescript, this pdf file is pasted as an attachment into the selected row.
Here is a screenshot how this can look like:
I am running this script via a key trigger in quicksilver, as described here. This works well, but is not nearly as nice as the beautiful icons that came with “kinkless gtd“, which later turned into omnifocus.
This is my reasoning for doing it:
- omnioutliner 4 has been released somewhat co-incidental with the recent mac birthday
- oo4 continues to have strong applescript support
- oo3eq is not around anymore; at least I couldn’t find it, and the version I still had around did not work anymore on my current setup
- the whole thing is reasonably quick, and I learned a few things about applescript and python
This post has to do neither with water nor with statistics. But because it’s been christmas break, I found some time to do some fun stuff with python:
I wrote a python script that downloads the issue of my choice of a pdf on my iPad. Partcularly, the script
- lets me chose which issue of the paper (relative to the current day) I want to download
- checks if that file might already exist on a relevant dropbox folder – checks if that file exists on the newspaper’s server (there is no paper on a holiday, f.ex) – if checks are ok, then the pdf is downloaded locally – saves the pdf in a selected folder within my dropbox – opens Goodreader, downloads the current file from dropbox (the proper url of the file needed to be determined) – deletes the downloaded pdf locally within pythonista
The rationale behind attempting this was two fold:
- using python on my iPad: I have downloaded pythonista a little while ago, but haven’t had time to play with it
since it has been possible, I have been subscribing to a German daily newspaper, which can be download as a pdf. I guess they are trying to foster the use of iPaper. I tried that for the last couple of weeks using iBooks on the iPad, and I don’t like it for two reasons:
- the iPaper files are larger than the pdfs, hence they take noticeably longer to load.
- I found it to be almost impossible to get the iPaper files out of iBooks on the iPad (to some storage place, for example).
The tipping bucket that led me to actually do this was that since a little while it is not possible anymore to download the pdfs directly from within Goodreader
Goodreader came out with an updated version two days ago. This makes the script even more powerful. This update solved issues related to the rendering of pdfs. I am not getting money from Goodreader, but I consider it to be the most useful software I have on my iPad. It particularly integrates awesomely with directories on an OSX Server.
Downloading the pdf
This is surprisingly easy using the
requests package, that I just found out that it existed. Basically, downloading a pdf from a known url comes down to one line:
r = requests.get(url, auth=('XXX', 'YYY'))
Saving to Dropbox
This is step is a little bit more involved:
- Before you start you have to register an ‘app’ in the developer part of dropbox. This will get you an
APP_SECRET. In the python API you also need a token, which is from what I understand a string consisting of the name of your app in dropbox concatenated with
- When you’ve got that, I found some code on github that deals with the authentication process.
Finally, within the python module within pythonista, saving a file comes down to another one liner basically (with some preparations):
sess = dropbox.session.DropboxSession(APP_KEY, APP_SECRET, ACCESS_TYPE) client = dropbox.client.DropboxClient(sess) f = open(filename) response = client.put_file('/Apps/Pythonista/' + filename, f)
- the link between pythonista and good reader via dropbox works reasonably well. Would be nice to have a direct link (along the lines of ‘open in’).
- it’s possible to create little home scree icons that call your code, also with command line arguments. As far as I can tell, you have to fix a certain value of the command line argument to one icon… that seems a bit against the flexibility of command line arguments
the time delta doesn’t treat boundaries of months correctlyfixed by the first commit
- not sure if
raise Exceptionis really the way to go if a) the file exists already on dropbox or b) if there doesn’t exist a paper on the desired date
- Pythonista offers a tool to achieve things that otherwise would be next to impossible on the iPad – at the same time, one has to wonder, why such things are forced to be impossible to begin with – writing code on the iPad without a dedicated external keyboard is… challenging. Despite the good editor of Pythonista.
The script is available on github.
I’ll start this year by pointing to two articles from the New York Times that point out the importance of uncertainty:
1. Monte Carlo to try and Find Fisherman who Fell off his Boat
This story in the New York Times Magazine describes the rescue for a fisherman who fell off his boat into the Atlantic while his colleagues were sleeping. The human aspects play the most important part, but one section describes how the US Coast Guard uses a Monte Carlo based simulation tool to predict the most likely locations where to search for the fisherman floating in the ocean. I found this story via John Gruber.
2. Inconvenient Uncertainties
In this comment, on another article published in the New York Times, Gernot Wagner tries to make the point that science in general, particularly environmental science is inherently uncertain . The title of the original article that he refers to is “By 2047, Coldest Years May Be Warmer Than Hottest in Past, Scientists Say.” Wagner tries to make the point, that science in general and climate science in particular, are uncertain. Will this threshold be passed exactly in 2047 or might it be around 2050? He continues to state that
The scientific method imposes some order, but in the case of climate change, that order is probabilistic. For the sake of science and the planet, we should not become distracted by a false sense of certitude. Imprecise truths are the most inconvenient ones. We know enough to act now. What we don’t know should prompt us to even more decisive action.