• Table of contents
• Cooper updates
  • 2014-5-13
  • 2013-10-22
  • 2013-09-24
  • 2013-08-13
  • 2013-06-04
  • 2013-05-21
  • 2012-12-17
  • 2012-07-17
  • 2012-06-28
  • 2012-05-21
  • 2012-05-08

Cooper updates¶

2014-5-13¶

To-Dos:
- Begin validations on derived variables and fitness tests on GEVs.
- Fix sunrise equation.

2013-10-22¶

What I did last week:
- Worked on climate extreme script.
- Cleaning up/documenting existing code.

What I'm working on now:
- Producing extreme summaries for 1980-2010 in Oregon.

2013-09-24¶

What I did last week:
- Researched sources of data for layer validation.
- Worked on the climate extreme scripts.
- Did some bug-fixing on the degree-day/ET script.

What I'm working on now:
- Producing Gumbel/Pareto/Poisson layers for Oregon.
- Wrapping up validation for derived variables.
- Drafting the final list of deliverables.

2013-08-13¶

What I did last week:
- Produced warming/cooling degree day layers.
- Produced basic climate extreme layers.
- Produced reference evapotranspiration layers.

What I'm working on now:
- Producing climate extreme value summaries.
- Validating / correcting preliminary layers.

2013-06-04¶

What I did last week:

- Finished the first draft of the DeriveIt script.
- Tested the first draft (actually, it's running right now - progress pending)

What I plan to do this week:

- Celebrate my birthday!
- Identify any runtime problems with the script.
- Finish the first draft of the ClimEx climate extreme script.

2013-05-21¶

What I did last week:

- Fixed some bugs in the DeriveIt derived variable script.
- Studied what kind of code/software already exists for extreme value theory modeling.

What I plan to do this week:

- Produce first draft layers using DeriveIt.
- Write up ClimEx climate extreme event model this weekend.
- Ideally, we should be seeing some layers created from ClimEx by next meeting.

2012-12-17¶

- Implemented vectorized calculations to improve scalability.
- Implemented a smarter data I/O so that it can take many single-band files or one multi-band file (as is the case with DAYMET).
- Testing derived climate/climate extreme scripts using DAYMET data.

_TO-DO

- Validation for Oregon: this will involve finding accurate PET/degree-day values. This could be challenging because it's not a common product.
- Validation of degree-day methodology: the hourly-temperature estimation curves were created empirically in North America. Since they're fitted to sunrise/sunset and tmax/tmin, it might not be a problem. But would be sensible to investigate adjustments based on location. This might require some additional modeling.

- Method

2012-07-17¶

What I did this past week

- Identified and fixed issues with degree-day algorithm in periods of unusual sunrise/sunset hours.
- Implementing scripts for bioclimatic variables already supplied by WorldClim (BIOCLIM variables).

What I'm working on now

- Getting Oregon layers created.

What's next

- Pending final version and feedback from Brian (this Friday).

2012-06-28¶

What I did this past week

- Finished the code.

What I'm working on now

- Documenting, commenting, and revising the script with Brian to make sure I'm on the right track.

What's next

- Incorporate Brian's feedback; make any necessary changes.
- Prepare some sample data layers for the next meeting.

2012-05-21¶

What I did this past week

- Finished the overall design of the derived climate package for Python.

What I'm working on now

- Debugging the code/cleaning it up.
- Studying NetCDF

What obstacles are blocking progress

- None so far.

What's next

- Send code to Brian McGill on Wednesday.
- Send code to everyone else after his proposed changes for even more feedback.
- Become comfortable with CDO/NCO tools by next meeting.

2012-05-08¶

What I did this past week

- Studied different methods of calculating radiation - CRU's radiation tiles, daily temperature amplitude, and GRASS' r.sun function.
- Closed a few knowledge gaps in Python and GRASS.
- Figured out raster algebra with tiles of different sizes.

What I'm working on now

- Adding command line functionality that will let me (and other users) use different estimation methods for a single derived variable.
- Working on functionality that calculates extreme values for user-defined intervals (seasonal extrema, warmest/coldest day in the past X years).

What obstacles are blocking progress

- CRU's cloud cover and vapor pressure tiles only seem to go up to 2003 and won't be useful for calculating derived variables for recent time-periods.

What's next

- Getting up to speed on github.
- Add some basic documentation and error-catching that will make the code more reusable.
- Submit said code for feedback!