Intro.

1
Intro.
Briefing for CCSP Observations Working Group May
10, 2004 Model-Observations Integration Randa
ll Dole, NOAA, CCSP Co-lead Climate Variability
and Change (CVC) and Climate Modeling (CM)
Working Groups

• Why is this integration essential to a climate
  observing strategy?
• Role in the CCSP Strategic Plan?
• CCSP Synthesis and Assessment Product
  Reanalysis/attribution.
• Science and technical challenges.
• Some key issues.

2
Why?
Why is this integration essential?

• Integrating diverse observations into a
  physically-based model through the process of
  data assimilation is vital for constructing near
  real-time climate analyses and periodic
  reanalyses of past climate.
• Modern data assimilation techniques enable data
  from many disparate observing systems to be
  effectively integrated together within a model to
  form a consistent climate system analysis.
• Climate analyses and reanalyses have numerous
  climate applications. They directly support
  research to advance understanding and
  predictions of climate, diagnose deficiencies in
  climate models, clarify observing system
  requirements and data set needs, and develop
  decision support resources.
• Climate analyses that are obtained by
  assimilating observations into a state-of-the art
  model are an essential component of an end-to-end
  climate observing system.

3
Climate analyses in the CCSP
Role in the CCSP
The fundamental need for ongoing, near real-time
climate analyses, together with periodically
updated climate reanalyses, is well recognized in
the CCSP SP. This need is not new. It has been
articulated repeatedly by the scientific
community and science advisory panels for over a
decade. Specific CCSP goals, questions,
objectives, research foci, and products directly
connected to or dependent on climate analyses are
listed in Ch 2. Integrating Climate and
Global Change Research Ch 4. Climate
Variability and Change. Ch 10. Modeling
Strategy. Ch 11. Decision Support Resources
Development. Ch 12. Observing and Monitoring the
Climate System. Ch 13 Data Management and
Information.
4
Why is climate analysis/reanalysis such a high
observational priority for CVC/CM WGs?

• Climate analyses derived from data assimilation
  integrate observational data with climate models.
  They can be used to evaluate model deficiencies
  and identify areas where improvements in models
  and observational data may be particularly
  beneficial. They are virtually essential for many
  diagnostic studies, as well as S-I prediction
  research.
• Ongoing climate analyses and updated reanalyses
  would support both CVC and CM needs more than any
  single observational data set.
• This integrating activity would help to address
  a broader array of questions and likely provide a
  higher return-on-investment than any single
  observational data set.
• Climate analysis and reanalysis products serve a
  very broad community, including scientists and
  end users.

5
CCSP Synthesis and Assessment Product

• Product Reanalysis of historical climate data
  for key atmospheric features. Implications for
  attribution of causes of observed change.
• Significance Understanding the magnitude of
  past climate variations is key to increasing
  confidence in the understanding of how and why
  climate has changed and how it may change in the
  future.
• Primary end use To inform policy decisions.
• Proposed lead agencies NOAA, NASA, DOE
  supporting.
• CCSP WGs Initially assigned to OWG, but CVC and
  CM WG have a strong interest and have been
  involved in early science planning.
• Time frame 2-4 years.

6
CCSP Reanalysis Synthesis ProductUpdate on
Progress

• Interagency Science Working Group has been
  meeting since December.
• Co-chairs Siegfried Schubert (NASA), Glenn
  White (NOAA).
• Approximately 20 participants, from NOAA, NASA,
  DOE, NCAR, University community.
• Draft plan developed for proposed products.

7
Proposed primary products

• Proposed Products
• A. State-of-science science reviews
• 1. Assessment of first generation reanalysis
  products
• 2. Assessment of current understanding of
  causes of 20th
• century climate variability and
  trends
• Develop high quality observational datasets.
• 1. For satellite era, correct sat.
  biases/trends, land surface, ocean data sets
  required for coupled data assimilation.
• 2. For pre-radiosonde era, QC surface press.
  obs crucial.
• C. Initiate next generation reanalyses
• Three proposed activity streams
• 1. Satellite era (1979 to present)
• 2. Period with substantial upper air
  network ( post-1948)
• 3. Period with minimum set of surface
  obs (1895 to present)

8
Proposed SARs

• Synthesis and Assessment Reports
• Proposed topics
• 1) State-of-science assessment of strengths and
  weaknesses of first generation reanalyses, their
  suitability and limitations for studies of
  climate variability and trends.
• State-of-science assessment of present knowledge
  of understanding and uncertainties of causes of
  observed climate variations and trends during the
  20th century.
• Possible additional report, or in 1)
• 3) Assessment of progress since first
  generation reanalysis to improve climate analyses
  and reanalyses, key issues, and necessary further
  steps to address outstanding science and
  policy-relevant questions.

9
Scientific and Technical Challenges

• Data inhomogeneities in space and time.
• Model biases.
• Optimizing analyses for climate purposes.
• Optimal use of data, minimization of spurious
  trends, bias, etc.
• Improved representation of processes and forcing,
  e.g., precipitation, clouds, interactions with
  surface.
• Better horizontal and vertical resolution.
• Estimating uncertainties.
• Use of data assimilation to extend reanalysis
  back in time.

10
Analysis of pre-radiosonde era
Example of use of modern data assimilation
methods to integrate observations with a
model. Feasibility of a pre-radiosonde era
reanalysis

• Current analyses for the pre-radiosonde (ca.
  1948) period consist of subjectively produced
  hand-drawn SLP maps that did not use all
  available observations. Can modern data
  assimilation systems be used to improve on these
  analyses? Can this approach provide us with
  additional information on the large-scale
  tropospheric anomalies, e.g., during the dust
  bowl years?
• A feasibility study was conducted using data
  removal and ensemble data assimilation
  techniques. Simulated reanalyses use only surface
  pressure observations at densities and temporal
  intervals representative of earlier years (1895,
  1915, 1935).

Major science/policy-relevant question Can we
better describe and interpret the causes of past
climate variability and change over the last
100-150 years?
11
500mb Height Analyses for 0Z 15 Dec 2001
5500 m contour is thickened Black dots show
pressure observation locations
Full CDAS (120Kobs)
EnSRF 1895 (214 surface pressure obs)
RMS 39.8 m r (z,NH) 0.96
OI 1895 (214 surface pressure obs)
RMS 82.4 m
12
Results indicate that

1. Reanalyses of the lower-tropospheric circulation
   prior to 1948 are feasible using just the
   available surface pressure observations.
2. Recent advances in ensemble data assimilation
   methods may lead to even better analyses,
   including for the upper troposphere.
3. Providing additional observations, especially in
   data sparse regions, will produce further
   improvement. Present approaches and data coverage
   should enable lower tropospheric reanalyses that
   are as accurate as current 2-3 day forecasts.
4. Having the most complete and carefully
   quality-controlled surface pressure data sets
   available will be especially crucial for
   historical reanalyses.

13
Some key issues

• Leadership. Which Agency(ies)? Which WG(s)?
• Short-term deliverables and data set development.
  What else could or should be accomplished before
  FY08? Needs differ for three streams of
  reanalyses, and for ongoing climate analyses.
• Support for data assimilation research,
  integration of observational and modeling
  capabilities.
• Roles and responsibilities. Coordination in CCSP,
  across agencies, and with extramural community.
• International coordination. Linkage to EOS/GEO.

14
Concluding comments

• A complete climate observing system requires both
  ongoing, near-real time climate analyses and
  periodic reanalyses using improved data sets and
  data assimilation methods. Both must be
  considered as essential components of a long-term
  climate observing strategy.
• So far, climate analyses and reanalyses have
  focused on the atmosphere. A longer-term strategy
  must be developed to analyze and eventually bring
  together the other, disparate components of the
  Earth system (oceans, land, cryosphere,
  hydrology, biosphere) through coupled model
  assimilation. This will enable a more
  comprehensive synthesis and understanding of the
  climate system.