/[MITgcm]/www.ecco-group.org/automatic.htm
ViewVC logotype

Diff of /www.ecco-group.org/automatic.htm

Parent Directory Parent Directory | Revision Log Revision Log | View Revision Graph Revision Graph | View Patch Patch

revision 1.2 by heimbach, Fri Jun 15 16:09:55 2007 UTC revision 1.3 by heimbach, Fri Oct 12 02:22:06 2007 UTC
# Line 93  atmospheric) state estimate to a wide co Line 93  atmospheric) state estimate to a wide co
93                <p class="headergray"> (AD) </p>                <p class="headergray"> (AD) </p>
94                Automatic Differentiation   (AD) is a technology for automatically augmenting computer programs, including   arbitrarily complex simulations, with statements for the computation of   derivatives (tangent linear, adjoint, Hessian, etc.), also known as   sensitivities. AD tools in our context provide source-to-source transformation   of a function, given as comouter code, to generate efficient and accurate   (truncation-free) code for computing first, second and higher-order derivatives   of the given function. <BR />                                  Automatic Differentiation   (AD) is a technology for automatically augmenting computer programs, including   arbitrarily complex simulations, with statements for the computation of   derivatives (tangent linear, adjoint, Hessian, etc.), also known as   sensitivities. AD tools in our context provide source-to-source transformation   of a function, given as comouter code, to generate efficient and accurate   (truncation-free) code for computing first, second and higher-order derivatives   of the given function. <BR />                  
95                <BR />                <BR />
96                  Since the mid-1990's, groups at MIT, SIO, JPL and   GFDL have applied AD tools for generating tangent linear and adjoint code for   ocean circulation and climate studies. The tools used comprised the <A href="http://www.autodiff.com/tamc" target="_blank">Tangent linear and Adjoint   Model Compiler (TAMC)</A> as well as its successor, <A href="http://fastopt.de/products/taf/taf.html" target="_blank">Transformation of   Algorithms in Fortran (TAF)</A>. This system has so far been used in a practical   way to study three broad classes of problems: (i) parameter sensitivity of the   climate system, (ii) initial and boundary value sensitivity, and (iii) ocean   state estimation (or data assimilation). A description of the implementation of   AD in the context of the <A href="http://mitgcm.org" target="_blank">MITgcm</A> is given in   chapter 5 of the <A href="http://mitgcm.org/r2_web_testing/latest/online_documents/node1.html" target="_blank">MITgcm Online Documentation</A>. <BR />                  Since the mid-1990's, groups at MIT, SIO, JPL and   GFDL have applied AD tools for generating tangent linear and adjoint code for   ocean circulation and climate studies. The tools used comprised the <A href="http://www.autodiff.com/tamc" target="_blank">Tangent linear and Adjoint   Model Compiler (TAMC)</A> as well as its successor, <A href="http://fastopt.de/products/taf/" target="_blank">Transformation of   Algorithms in Fortran (TAF)</A>. This system has so far been used in a practical   way to study three broad classes of problems: (i) parameter sensitivity of the   climate system, (ii) initial and boundary value sensitivity, and (iii) ocean   state estimation (or data assimilation). A description of the implementation of   AD in the context of the <A href="http://mitgcm.org" target="_blank">MITgcm</A> is given in   chapter 5 of the <A href="http://mitgcm.org/r2_web_testing/latest/online_documents/node1.html" target="_blank">MITgcm Online Documentation</A>. <BR />
97                    <BR />                    <BR />
98                  The current ocean state   estimation effort may be considered to be among the most complex inverse   modeling exercises attempted to date. It has placed the <A href="#">ECCO consortium </A>into a position to push the   limits of AD, and given it experience and expertise to address questions as to   where and how AD needs to be improved for large scale applications. <BR />                  The current ocean state   estimation effort may be considered to be among the most complex inverse   modeling exercises attempted to date. It has placed the <A href="#">ECCO consortium </A>into a position to push the   limits of AD, and given it experience and expertise to address questions as to   where and how AD needs to be improved for large scale applications. <BR />
99                    <BR />                    <BR />
# Line 107  atmospheric) state estimate to a wide co Line 107  atmospheric) state estimate to a wide co
107                <UL>                <UL>
108                  <LI><A href="http://www.autodiff.org" target="_blank">Autodiff.org</A>                  <LI><A href="http://www.autodiff.org" target="_blank">Autodiff.org</A>
109                  </li><LI><A href="http://www.autodiff.com/tamc/" target="_blank">TAMC</A>                  </li><LI><A href="http://www.autodiff.com/tamc/" target="_blank">TAMC</A>
110                  </li><LI><A href="http://www.fastopt.de/taf/" target="_blank">TAF</A>                  </li><LI><A href="http://fastopt.de/products/taf/" target="_blank">TAF</A>
111                  </li><LI><A href="http://www-unix.mcs.anl.gov/autodiff/ADIFOR/" target="_blank">ADIFOR</A>                  </li><LI><A href="http://www-unix.mcs.anl.gov/autodiff/ADIFOR/" target="_blank">ADIFOR</A>
112                  </li><LI><A href="http://www-fp.mcs.anl.gov/adic/" target="_blank">ADIC</A>                  </li><LI><A href="http://www-fp.mcs.anl.gov/adic/" target="_blank">ADIC</A>
113                  </li><LI><A href="http://tapenade.inria.fr:8080/tapenade/index.jsp" target="_blank">TAPENADE</A>                  </li><LI><A href="http://tapenade.inria.fr:8080/tapenade/index.jsp" target="_blank">TAPENADE</A>
Legend:
Removed from v.1.2  
changed lines
  Added in v.1.3
  ViewVC Help
Powered by ViewVC 1.1.22