|
Size: 2818
Comment:
|
Size: 7743
Comment:
|
| Deletions are marked like this. | Additions are marked like this. |
| Line 1: | Line 1: |
| #acl AdminGroup:read,write,delete,revert EditorGroup:read,write StudentGroup:read,write | #acl AdminGroup:read,write,delete,revert EditorGroup:read,write All:read |
| Line 5: | Line 5: |
| Line 8: | Line 7: |
| = 63-953 Climate and Satellite Data Analysis = | = 63-953 Climate and Satellite Data Analysis = '''Lars Kaleschke, Alexander Loew''' |
| Line 14: | Line 14: |
| Place: Geom 1536c | Place: Geom 1536c |
| Line 17: | Line 17: |
The participants will learn to practically work with climate model, reanalysis, in-situ station and satellite data. Organized as a group project, the participants will further learn the principles of project management and shared software development. |
The participants will learn to practically work with climate model, reanalysis, in-situ station and satellite data. Organized as a group project, the participants will further learn the principles of project management and shared software development. |
| Line 21: | Line 20: |
| === Monday === |
=== Monday, 10:00 === |
| Line 25: | Line 23: |
| * [[/PM|Project management and documentation]] | * [[Climate_and_Satellite_Data_Analysis_2014/PM|Project management and documentation]] |
| Line 27: | Line 25: |
| * [[DataIO|Access to data sets]] * [[/Plotting|Plotting]] [[attachment:Seawater plot.ipynb]] * [[GenMaps|Data Visualization: Generating maps]] * [[http://ipython.org/ipython-doc/dev/interactive/htmlnotebook.html|IPython Notebooks]] [[https://github.com/ipython/ipython/wiki/A-gallery-of-interesting-IPython-Notebooks|Gallery of interesting IPython Notebooks]] * [[attachment:Stochastic analysis of time series.ipynb]] [[attachment:Monte carlo.ipynb]] [[attachment:GISSTEMP netcdf.ipynb]] [[attachment:datetimeobjects.ipynb]] [[attachment:Station data seasonal cycle.ipynb]] [[/Projects2014 | Projects descriptions]] |
* [[DataIO|Access to data sets]] * [[GenMaps|Data Visualization: Generating maps]] |
| Line 42: | Line 35: |
| Morning: Group presentations of project plan and preliminary analysis. | |
| Line 43: | Line 37: |
| Morning: Group presentations of project plan and preliminary analysis. | |
| Line 47: | Line 40: |
| Morning: Group presentations of methods and code implementations | |
| Line 48: | Line 42: |
| Morning: Group presentations of methods and code implementations | |
| Line 52: | Line 45: |
| Morning: Group presentations of preliminary results | |
| Line 53: | Line 47: |
| Morning: Group presentations of preliminary results | |
| Line 57: | Line 50: |
| Morning: final presenation of results and discussion | |
| Line 58: | Line 52: |
| Morning: final presenation of results and discussion Afternoon: evaluation and preparation of final report |
Afternoon: evaluation and preparation of final report |
| Line 61: | Line 54: |
| = Topics for group work = | |
| Line 62: | Line 56: |
| == Topics for group work == = Project groups = |
The major topic of this course will deal with data gaps in climate and satellite records and their implications for the calculation of global statistics. The overall objectives of the course are: * raise awareness on the problem with data gaps in climate records * learn about methods for coping with data gaps * assess the effect of data gaps on statistics of climate related datasets * learn processing of data using a variety of methods Another major objective of this course is to train the so called soft-skills in practice, like * project management (planning ... final report) * joint code development * reporting == Has the Earth stopped warming? == === Coverage bias in the HadCRUT4 temperature series and its impact on recent temperature trends === {{attachment:media_summary.png}} Kevin Cowtan and Robert Way published [[http://onlinelibrary.wiley.com/doi/10.1002/qj.2297/abstract | a paper]] which was quite controverily discussed in the scientific community. They fill gaps of the HadCRUT temperature data set by using satellite data. By filleing this (well known) Arctic gap and compare their new reconstruction of surface temperature data to independent in-situ observations and reanalysis data they show that the global mean temperature hiatus is not observable any more. Cowtan and Way (2013) methods and data are freely available and we will use them in the course: * http://onlinelibrary.wiley.com/doi/10.1002/qj.2297/abstract * http://www-users.york.ac.uk/~kdc3/papers/coverage2013/methods.html In situ measurements of surface temperatures are available from the International Arctic Buoy Programme (IABP) website: * http://iabp.apl.washington.edu/data_satemp.html === Project A: Cowtan and Way (CW2013) reconstruction === * Review methods of Cowtan and Way (2013) and give a summary of their approach * Analyse their reconstructed dataset of surface air temperature in terms of different parameters (global mean and stdv, seasonal cycle, anomalies, trends etc.) * is the dataset homogeneous in time or can you observe temporal inconsistencies? * how robust are trends estimated from this data record? What is the role of the length of the timeperiod? * do the same analysis for the original HadCRU temperature dataset (and possibly also for other datasets, like reanalysis (cross-check with Project C !)) * are results different? are they significantly different? * Look at correlations with climate indices, e.g. ENSO, NAO, PDO * Present all results in proper form (maps, statistics ...) |
| Line 68: | Line 98: |
| === Project B: Variations in Surface Air Temperature Observations in the Arctic === * Review methods of Rigor et al. (2000) * Analyse the surface air temperature (monthly mean gridded products) measured by the drift buoys * Calculate seasonal cycle, anomalies, and linear trends * Assess the significance of trends * Estimate the spatial correlation of surface temperature anomalies * Present all results in proper form (timeseries, maps, statistics ...) * [[http://iabp.apl.washington.edu/AirT/RigorEtal-SAT.pdf|Rigor, I., R. Colony, and S. Martin, 2000, Variations in Surface Air Temperature Observations in the Arctic, 1979 - 1997, J. Climate, Vol. 13, no 5, 896-914.]] |
|
| Line 69: | Line 109: |
=== Project C: Data intercomparison === Use the buoy measurements of surface air temperature as ground truth * Write code to interpolate the different datasets in a common grid * Compare CW2013, buoy and reanalysis data * Was the data gap in the Arctic filled in reasonably? * Are there biases or jumps in the data? |
|
| Line 72: | Line 121: |
| == Project D: HOAPS ocean flux sampling bias == The Meteorological Institute of the University of Hamburg and the Max-Planck-Institute for Meteorology have compiled a climate data record of ocean surface fluxes. This so called [[http://www.hoaps.org|HOAPS climatology]] has been sucessfully applied in numerous studies and is one of a very few global records on ocean surface fluxes. HOAPS is uniquely built based on satellite data. Major references for HOAPS can be found on the [[http://www.hoaps.org | project website]]. Details of the algorithms and data processing are provided in [[http://www.earth-syst-sci-data.net/2/215/2010/ | Andersson et al. (2010)]] and [[http://journals.ametsoc.org/doi/abs/10.1175/2010JAMC2341.1 | Andersson et al (2011)]]. The HOAPS dataset is based on sampling twice a day, which is due to the overpass time of the satellites used to generate the product. The project shall address the following research questions: * What is the impact of undersampling the diurnal cycle? * are biases introduced in e.g. monthly means? * what is the impact of sea ice gaps on monthly means? * What is the effec of different land/sea masks and spatial grids (resolution, projections) on total mean global fluxes? * do total mean global numbers change? If so, by how much? * how do global mean flux estimats from HOAPS compare to other datasets available? * How are HOAPS surface fluxes related to climate indices like e.g. NAO, PDO, ENSO ... ? * as a starting point the paper from [[http://tellusa.net/index.php/tellusa/article/view/15710 | Andersson et al (2010) ]] might be usefull * develop appropriate metrices (e.g. different correlation approaches, EOF) for comparing HOAPS surface fluxes with climate indices [[/ProjectD]] |
|
| Line 75: | Line 146: |
* prepare [[ftp://ftp.ifremer.fr/ifremer/cersat/products/gridded/psi-concentration/data/|sea ice data]] (Lars) :-) (CMIP5 data available) * prepare CMIP5 data (Alex) :-) |
* CW2013 :-) |
| Line 80: | Line 149: |
| * Stations :-) | * Bouy data :-) * HOAPS data :-) |
| Line 83: | Line 153: |
| Line 88: | Line 157: |
| * Methods and data | * Methods and data |
| Line 91: | Line 160: |
| * Conlcusion | * Conclusion |
| Line 94: | Line 163: |
| Line 97: | Line 165: |
= Examples from the past = == How significant are observations of Arctic temperature trends? == * https://wiki.zmaw.de/lehre/Climate_and_Satellite_Data_Analysis_2013/ProjectA |
Contents
63-953 Climate and Satellite Data Analysis
Lars Kaleschke, Alexander Loew
MS Integrated Climate System Sciences
Date: 3.2.2014-7.2.2014
Place: Geom 1536c
Course objectives
The participants will learn to practically work with climate model, reanalysis, in-situ station and satellite data. Organized as a group project, the participants will further learn the principles of project management and shared software development.
Schedule
Monday, 10:00
General Introduction
Group work: develop a project plan and write a short technical proposal for your project.
Final report due by 15. March 2014
Obtain data and do preliminary analysis (e.g. data coverage).
Tuesday
Morning: Group presentations of project plan and preliminary analysis.
Afternoon: implementation and project work
Wednesday
Morning: Group presentations of methods and code implementations
Afternoon: Project work
Thursday
Morning: Group presentations of preliminary results
Afternoon: Project work
Friday
Morning: final presenation of results and discussion
Afternoon: evaluation and preparation of final report
Topics for group work
The major topic of this course will deal with data gaps in climate and satellite records and their implications for the calculation of global statistics. The overall objectives of the course are:
- raise awareness on the problem with data gaps in climate records
- learn about methods for coping with data gaps
- assess the effect of data gaps on statistics of climate related datasets
- learn processing of data using a variety of methods
Another major objective of this course is to train the so called soft-skills in practice, like
- project management (planning ... final report)
- joint code development
- reporting
Has the Earth stopped warming?
Coverage bias in the HadCRUT4 temperature series and its impact on recent temperature trends
Kevin Cowtan and Robert Way published a paper which was quite controverily discussed in the scientific community. They fill gaps of the HadCRUT temperature data set by using satellite data. By filleing this (well known) Arctic gap and compare their new reconstruction of surface temperature data to independent in-situ observations and reanalysis data they show that the global mean temperature hiatus is not observable any more.
Cowtan and Way (2013) methods and data are freely available and we will use them in the course:
In situ measurements of surface temperatures are available from the International Arctic Buoy Programme (IABP) website:
Project A: Cowtan and Way (CW2013) reconstruction
- Review methods of Cowtan and Way (2013) and give a summary of their approach
- Analyse their reconstructed dataset of surface air temperature in terms of different parameters (global mean and stdv, seasonal cycle, anomalies, trends etc.)
- is the dataset homogeneous in time or can you observe temporal inconsistencies?
- how robust are trends estimated from this data record? What is the role of the length of the timeperiod?
- do the same analysis for the original HadCRU temperature dataset (and possibly also for other datasets, like reanalysis (cross-check with Project C !))
- are results different? are they significantly different?
- Look at correlations with climate indices, e.g. ENSO, NAO, PDO
- Present all results in proper form (maps, statistics ...)
Project B: Variations in Surface Air Temperature Observations in the Arctic
- Review methods of Rigor et al. (2000)
- Analyse the surface air temperature (monthly mean gridded products) measured by the drift buoys
- Calculate seasonal cycle, anomalies, and linear trends
- Assess the significance of trends
- Estimate the spatial correlation of surface temperature anomalies
- Present all results in proper form (timeseries, maps, statistics ...)
Project C: Data intercomparison
Use the buoy measurements of surface air temperature as ground truth
- Write code to interpolate the different datasets in a common grid
- Compare CW2013, buoy and reanalysis data
- Was the data gap in the Arctic filled in reasonably?
- Are there biases or jumps in the data?
Project D: HOAPS ocean flux sampling bias
The Meteorological Institute of the University of Hamburg and the Max-Planck-Institute for Meteorology have compiled a climate data record of ocean surface fluxes. This so called HOAPS climatology has been sucessfully applied in numerous studies and is one of a very few global records on ocean surface fluxes. HOAPS is uniquely built based on satellite data.
Major references for HOAPS can be found on the project website. Details of the algorithms and data processing are provided in Andersson et al. (2010) and Andersson et al (2011).
The HOAPS dataset is based on sampling twice a day, which is due to the overpass time of the satellites used to generate the product.
The project shall address the following research questions:
* What is the impact of undersampling the diurnal cycle?
- are biases introduced in e.g. monthly means?
- what is the impact of sea ice gaps on monthly means?
* What is the effec of different land/sea masks and spatial grids (resolution, projections) on total mean global fluxes?
- do total mean global numbers change? If so, by how much?
- how do global mean flux estimats from HOAPS compare to other datasets available?
* How are HOAPS surface fluxes related to climate indices like e.g. NAO, PDO, ENSO ... ?
as a starting point the paper from Andersson et al (2010) might be usefull
- develop appropriate metrices (e.g. different correlation approaches, EOF) for comparing HOAPS surface fluxes with climate indices
TODOs
Data
CW2013
ERA-Interim
NCEP
Bouy data
HOAPS data
Final report
Template structure:
- Abstract
- Introduction: state of the art (literature), statement of the problem
- Methods and data
- Results
- Discussion
- Conclusion
References
- Python Scripting for Computational Science, Hans Petter Langtangen, Springer (available in the ZMAW library)
Examples from the past