April 28, 2018

New Individual Tree Change Paper Online!

March 7, 2017

First meeting of CEOS LPV biomass co-leads in London at UCL (myself, Mat Disney and John Armston). Stay tuned for updates from this group, and if you would like to be added to the mailing list please let me know!


Aug 5, 2016

Happy to announce that our NASA CMS project, ‘Future Mission Fusion for High Biomass Carbon Accounting’ has been selected! Lola Fatoyinbo is the PI, I am the science PI, and together we will quantify the expected accuracy of NISAR, ICESAT-2, and GEDI in three high biomass forests. Our study sites will be in Sonoma County, California, Gabon and Costa Rica. Very excited to get this work under way!


New Scientific American article today – please note that GEDI is a University of Maryland mission led by my former PhD supervisor Ralph Dubayah, and lidar can penetrate TO the forest floor, not into it! But great press for GEDI.


Feb 26, 2016

Fun in the Gabonese jungle:

June 9, 2014

New RSE article online today looking at the utility of individual tree metrics vs. the scaling-based approaches presented by Greg Asner & Joe Mascaro:

May 7, 2013

Very inspired by the Dalai Lama’s talk at University of Maryland today. Perhaps we should not separate morality from science, but find new ways of integrating the two for the betterment of society. Watch the lecture here:

Also, totally unrelated, I finally found a great documentation of the Forest Inventory Analysis (FIA) data that I’m currently sifting through and writing scripts to pull out national DBH–> height relationships.

April 27, 2013

NEX Summer short course video released. It really was an awesome experience and I continue to use the supercomputer system almost daily:

Dec 15, 2012

End of a busy fall – AGU 2012 was a fantastic meeting both for our research group (plenty of great talks and posters) and as a whole (great talks from Mars, inspiring screening of Chasing Ice). I also defended my dissertation proposal yesterday to a packed room – thanks to all who attended. It’s great to be moving into a period of 100% research time!

I presented a poster outlining the importance of individual tree metrics for biomass modeling. The results so far show that the tree metrics give a slight improvement on model performance, but speculate that general models across different ecosystems may be even better.

September 6, 2012

Last minute prep on work before going to ForestSAT in Oregon followed by Silvilaser in Vancouver. I’ll be presenting some results from my delineation algorithm as a poster in Vancouver, and a comparison between tree size distribution and age since disturbance (from VCT) as a talk in Oregon.

July 29, 2012

Thoughts: NEX (NASA Earth Exchange) is an effort out of NASA Ames to combat pressing global issues by combining super computing capabilities, big data access and scientific collaboration. I was really privileged to be one of 8 students invited to attend the first NEX summer short course. The course was incredibly well organized by Rama Nemani and Jennifer Dungan, and featured a panel of brilliant and world renowned keynote speakers through week one, with hands on project work during week two. I learned how to parallelize my canopy delineation algorithm and it is now up and running on NASA’s Pleaides, the world’s 11th fastest super computer (4th in the US).

I think NEX represents the next generation of scientific enquiry in Earth Sciences – the data is too big and the problems too broad; we need to get better at team science, communication, and data management, and I think the NEX framework is a great start.

Tips:  For parallelizing IDL script:

– Spend $ on extra licenses (you need one per node), or buy MPIDL which apparently allows within IDL parallelization. I didn’t do this 🙂 Otherwise, translate your code into GDL (free version of IDL) but note that there are some small disparities. For example, reading ASCII csv files in GDL is slightly different than IDL – if you use the typical ascii IDL reading format:

data=fltarr(cols, file_lines(filename))

openr, lun, filename, /getlun

readf, lun, data

free_lun, lun

For a CSV file in IDL this will read the data in properly. In GDL it will only read into the first column of ‘data’, because it’s seeing the entire line of input as a single line and it will simply read into the first column. I hacked around this by converting my lidar data to tab delimited. Might be more useful to import as binary for other applications.

July 14, 2012

Getting excited to attend NASA Earth Exchange Supercomputing short course at NASA Ames – flying to San Francisco Sunday! You can remotely attend the conference here: