Appendix E

Climate Modeling Survey: Summary Responses

42 Responses Received

Note: As a point of reference, there were two unique questionnaires that were sent out to U.S. modeling centers for the purposes of this report. One questionnaire was sent out to large and intermediate centers, and a second questionnaire was sent to small centers. 1 Thus, the ‘coding' after each question, e.g., I6L (large/intermediate), I6S (small), specifies the question number as in the surveys above and whether it was common to both questionnaires, or exclusive to one or the other. In some instances, a question was specific to only one survey as it was believed to be inappropriate to the other category of modeling centers.

  1. What percentage of your modeling activities are devoted to operational versus research purposes? (I6L, I6S)

39 Majority research oriented

3 Majority operations oriented

  • Out of the responses that were majority research oriented, some stated that their research had direct operational relevance.

1  

An example of what is referred to in this document as a small modeling effort is one using a global, stand-alone atmospheric climate model at R15 (~4.5° × 7.5°) resolution; an example of an intermediate effort is one using a global, stand-alone atmospheric climate model at T42 (2.8° × 2.8°) resolution; an example of a large or high-end modeling effort is one using a global, coupled T42 atmospheric / 2° × 2° oceanic model (or finer resolution) for centennial-scale simulations of transient climate change.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 105
Improving the Effectiveness of U.S. Climate Modeling Appendix E Climate Modeling Survey: Summary Responses 42 Responses Received Note: As a point of reference, there were two unique questionnaires that were sent out to U.S. modeling centers for the purposes of this report. One questionnaire was sent out to large and intermediate centers, and a second questionnaire was sent to small centers. 1 Thus, the ‘coding' after each question, e.g., I6L (large/intermediate), I6S (small), specifies the question number as in the surveys above and whether it was common to both questionnaires, or exclusive to one or the other. In some instances, a question was specific to only one survey as it was believed to be inappropriate to the other category of modeling centers. What percentage of your modeling activities are devoted to operational versus research purposes? (I6L, I6S) 39 Majority research oriented 3 Majority operations oriented Out of the responses that were majority research oriented, some stated that their research had direct operational relevance. 1   An example of what is referred to in this document as a small modeling effort is one using a global, stand-alone atmospheric climate model at R15 (~4.5° × 7.5°) resolution; an example of an intermediate effort is one using a global, stand-alone atmospheric climate model at T42 (2.8° × 2.8°) resolution; an example of a large or high-end modeling effort is one using a global, coupled T42 atmospheric / 2° × 2° oceanic model (or finer resolution) for centennial-scale simulations of transient climate change.

OCR for page 105
Improving the Effectiveness of U.S. Climate Modeling Please describe groups with which you have significant collaboration and briefly describe the nature of this collaboration. (I7L, II1S) From the responses received, there appears to be strong connections between the major centers and academia and vice-versa. Please provide your opinion on current U.S. climate and weather modeling capabilities relative to overseas efforts. Please describe where differences in capabilities exist and what you feel are the causes for these differences. (II1L, III1S) U.S. is:     Ahead Behind Comparable Weather 2 20 6 Climate 1 21 9 Why are there differences? Underfunded Understaffed Lack of computer resources Lack of common center/coordination Other statements: Comparable to other countries at all but high-end Model development is weak here and overseas U.S. is ahead in diversity and size of effort It is more difficult to organize the U.S. effort due to its size and diversity If you stated that U.S. climate and weather modeling capabilities are behind those of other countries, do you have any suggestions to remedy this deficiency? (II2L, III2S) 7 Increased Funding 8 Shared Infrastructure 18 Enhanced Organization 25 Hardware 8 Adequate brainpower Do you feel that your modeling effort would be aided by altering the organization of U.S. climate modeling resources? If so, what changes would you recommend be made? (II3S) 6 Yes 5 No

OCR for page 105
Improving the Effectiveness of U.S. Climate Modeling Observations of the affirmative responses: Too many underfunded, understaffed groups Inadequate links to data collection More emphasis should be placed on the vulnerability of the Earth system to the spectrum of environmental stresses, rather than focus primarily on the effects of greenhouse gases. U.S. should take the lead in the physics of the climate system and its parameterization Devolve computing resources away from computer centers to the users Develop a responsive, interactive computing environment Make it easier to access climate models for climate applications and to long-term model simulation data for analysis What additional upgrades would be incorporated if funds were available? (III2bL, V2S) 7 Upgrades for PC clusters 2 More nodes 3 Increased bandwith 7 Increase general computational power 5 Increase disk storage 3 Increase file migration capabilities 1 Purchase Alpha-type workstations 1 Upgrade to parallel vector systems if possible 2 None 6 More processors Does modeling capacity or capability limit your current activities or does some other factor? Could you make use of additional modeling capacity or capability for additional activities? (III3bL, V3S) 27 Yes 2 No 7 Additional human resources 18 Additional computing capabilities Is computational time shared with the wider community? If so, how is this interaction organized? (III7L) 9 Yes 12 No 2 Yes, via scientific collaboration

OCR for page 105
Improving the Effectiveness of U.S. Climate Modeling 1 Only within DOD and with DOD funded scientists 3 Sharing is through a queuing system 2 Sharing through proposals for computer time 1 Sharing only within DOE 3 Via allocation process 1 Via output only Please provide your thoughts on the relative merits and hindrances of running your models on massively parallel processing sys tems relative to parallel vector architectures. (IV1L) 4 Massively parallel architecture is better 18 Parallel vector architecture is better MPP architecture is better but: There are a lack of compilers for these systems The transfer of code to MPP is not easy Vendors are not ready to supply the needed systems Other comments: MPP is harder to use MPP benefit is that the processing time is cheaper as the cost of the systems and maintenance is less than for parallel vector systems MPP offers more CPU power and memory per dollar spent Some new models can only be run on MPP MPP requires longer code development MPP is not scalable MPP offers poor system software and is unstable MPP requires additional personnel MPP offers poor communication among processors Do you use models or outputs from other facilities? If so, are any restrictions placed on the models or data? (VI3S) 12 Yes 0 No Restrictions: Output is restricted to research collaborators DOE security restrictions on computing access No restrictions Some foreign data is restricted Some data are restricted due to being in a pre-release state

OCR for page 105
Improving the Effectiveness of U.S. Climate Modeling How “portable” is your code without experiencing major performance loss? (IV6L) 22 Very portable 2 Somewhat portable 8 Code is custom altered for specific platforms 1 Code works on MPP only 1 Code is portable to VPP and MPP with some limitation Are you currently planning to (or intending to in the future) convert model codes to run on massively parallel machines? If cur rently converting, what experience do you have with this process? If intending to in the future, what are your plans for doing so? (V5L) 10 Already converted 12 Underway 2 Not underway Are model results produced by your facility made available to the wider scientific community? If so, are any restrictions placed on the models or data? (VI2S) 10 Yes 0 No Are model results produced by your facility made available to a wider scientific community? If so, are any restrictions placed on the data? (IV2L) 3 Yes with some restrictions 21 Yes 1 Yes, but only with collaborators 0 No Additional: 2 More widely distributed if resources were available 1 Yes, through published work Are the number of staff supporting your efforts sufficient? If not, please describe where improvements are needed. (V2L) 6 Yes 20 No Staff needed for:

OCR for page 105
Improving the Effectiveness of U.S. Climate Modeling Data interpretation and analysis Programmers Software engineers Hardware maintenance Model simulation interpretation High-performance applications Are the number of staff supporting your efforts sufficient? If not, please describe where improvements are needed. (VII2S) 4 Yes 8 No Staff needed for: Data interpretation and analysis Programmers Model developers Do you feel that your efforts are being limited by access to high-end computing resources? By access to model output from large modeling centers? By availability of diagnostic tools? By any other factors? (VIII1S) 11 Yes 1 No 1 Skilled personnel are not centrally located 1 No long-term strategy 1 Data outputs need to be made more user friendly 1 Satellite data needs to be made more user friendly 9 Access to computing 1 Access to global models 1 Stable funding Do you feel that your modeling efforts are being limited by lack of sufficient high-end computing resources? By people? By other resources? By any other factors? (VI1L) 26 Yes 27 No Factors: 17 People 18 Computing Other factors:

OCR for page 105
Improving the Effectiveness of U.S. Climate Modeling Lack of well-documented modern model codes Network bandwidth Data storage Stable funding When staffing positions in the categories listed above, what are the main difficulties, if any, involved (i.e. level of training required, salary requirements). (V3L, VII3S) 1 Research is very specialized 16 Salary is not competitive 10 Finding funding 15 Level of training 7 Difficult to find qualified programmers 1 Navy bureaucracy 3 Difficult to find model developers 1 No difficulty Please describe any future changes in staffing that are planned. (V4L) 7 None 8 Model/software support 8 Scientist 5 Modeler 1 Hardware Please describe any future changes in staffing that are planned. (VI4S) 4 None 4 Model/software support 3 Scientist 0 Hardward maintenance What is your highest priority if some of these limiting factors are removed? (VI2L) 11 Enhanced computing capabilities 8 Enhanced human resources 7 Improved physical performance of the models 1 Build a modeling system infrastructure 4 Increase the number of models 7 Increase model resolution

OCR for page 105
Improving the Effectiveness of U.S. Climate Modeling 1 Develop a high performance regional climate model 1 Adapt model code for parallel systems 1 Perform simulations on non-local systems 1 Additional R& D research funding Do you feel that future modeling efforts will be hindered by the availability of quality graduate students? If so, what steps would you recommend to remedy this problem? (VI5S) 3 No 5 Yes