Environmental applications research



Download 365.21 Kb.
Page1/13
Date conversion09.01.2018
Size365.21 Kb.
  1   2   3   4   5   6   7   8   9   ...   13

ENVIRONMENTAL APPLICATIONS RESEARCH

Principal Investigator: T.H. Vonder Haar

NOAA Project Goals: Various
Keywords: Various
1. Long-term Research Objectives and Specific Plans to Achieve Them:
Various. See following reports.
2. Research Accomplishments/Highlights
See following reports.
3. Comparison of Objectives Vs. Actual Accomplishments for Reporting Period:
See following reports.
4. Leveraging/Payoff: See following reports.
5. Research Linkages/Partnerships/Collaborators, Communication and Networking:
See following reports
6. Awards/Honors: See following reports.
7. Outreach: See following reports
8. Publications: See following reports



I. Research Collaborations with the ESRL/GSD Office of the Director



Project Title:The Use of Unmanned Aerial Systems for Atmospheric Observations
Principal Researcher: Nikki Prive'
NOAA Project Goal / Program: Climate—Understand climate variability and change to enhance society's ability to plan and respond / Climate observations and analysis
Key Words: Unmanned Aerial Systems (UAS), Climate, Observations, Arctic

1. Long-term Research Objectives and Specific Plans to Achieve Them:


To develop a concept of operations for a global network of Unmanned Aerial Systems (UAS) for the purpose of improving atmospheric observations for climate and weather over data-poor regions. The observational goals and viability of such a network will be determined by using Observing System Simulation Experiments (OSSE) for contributions to operational weather forecasting; and through analysis of existing climate data for contributions to climate research. Numerical modeling and analysis of past data will be used to determine the optimal choices of UAS routing.
2. Research Accomplishments/Highlights:

Code for the generation of synthetic UAS observations were debugged and ported to the wjet computer system. Access to the NCEP supercomputer was obtained and an account for the GSI data assimilation package has been set up. A method of converting the synthetic UAS observations into BUFR format for ingestion into the GSI was developed and testing of the assimilation of the synthetic data into the GSI is currently underway. Initial testing of the OSSE system is confined to the NCEP supercomputer at the moment due to unavailability of software (GFS/GSI) on wjet. Error characteristics for the synthetic observations have not been included in initial testing of the OSSE system, and are expected to be addressed during the next stage of OSSE development (calibration).

Evaluation of the Nature Run for the Pacific UAS testbed OSSE (Rossby waves) and Arctic Testbed OSSE were performed. The Rossby wave dynamics in the Nature Run were compared with wave behavior in the ECMWF and NCEP reanalysis; Arctic lows were similarly investigated. The boundary layer physics and vertical profile of the Arctic were also compared with observations from sondes, surface data, and upper air observations.
3. Comparison of Objectives vs Actual Accomplishments for the Report Period:
--Objective: Running an observing system simulation of the effect of large numbers of UAS sondes on the medium range weather forecast.
Status: In progress. Code for the generation of synthetic UAS observations was completed. Diagnostic evaluation of the Nature Run is ongoing.
--Objective: Support for the development of an Arctic UAS testbed.
Status: On-going. A panel of UAS testbed leads has been formed and awaiting their determination of further projects for support of the three proposed testbeds in the Arctic, Pacific, and Gulf regions.
4. Leveraging / Payoff:
The proposed network of UAS observations, envisioned to be a key component of the NOAA-proposed GEOSS, would provide regular vertical profiles of atmospheric conditions across data-poor regions, with the goal of improving operational weather forecasting and providing quality data for climate change research. The current efforts to design and optimize the proposed UAS observational network help to ensure that the network would be viable and successful at reaching these goals.
5. Research Linkages/Partnerships/Collaborators, Communications and Networking:
6. Awards/Honors:
7. Outreach:
8. Publications:
Project Title: Flow-following Finite-volume Icosahedral Model (FIM)
Participating CIRA Researcher: Brian Jamison

NOAA Project Goal / Program: Weather and Water—Serve society’s needs for weather and water information / Environmental modeling and Climate—Understand climate variability and change to enhance society's ability to plan and respond / Climate predictions and projections

Key Words: Flow-following, Finite-volume Icosahedral Model
1. Long-Term Research Objectives and Specific Plans to Achieve Them:
The FIM model is a new global model that is being developed at GSD. It features an isentropic-sigma hybrid vertical coordinate system (successfully used in the Rapid Update Cycle (RUC) model) and an icosahedral horizontal grid (Fig. 1). Tasks for this project include: generating graphics of output fields, creation and management of websites for display of those graphics, and creation and management of graphics for hallway public displays, including software for automatic real-time updates.

Fig. 1. An example of the icosahedral grid, with FIM model temperature plotted (provided by Ning Wang).

2. Research Accomplishments/Highlights:


Interpolation routines were developed that generate FIM output on a 0.5 degree latitude and longitude grid. These output fields can then be plotted using standard contouring packages and can be compared with other global models such as the Global Forecast System (GFS) model. Currently, a cylindrical equidistant projection is used (Fig. 2); however, other projections are being investigated.


Fig. 2. A cylindrical equidistant plot of surface "skin" temperature from a 12-hour FIM forecast interpolated to a 0.5 degree latitude-longitude grid.

A website for display of FIM model output has been created and has some preliminary products available for perusal with 3-hourly forecasts going out to 7 days (http://fim.noaa.gov/fimgfs). The website is currently being enhanced to include more output products, GFS model output, and FIM-GFS difference fields.

A dual-monitor hallway display was installed on the second floor of the David Skaggs Research Center (DSRC) to display FIM model graphics for public viewing. Currently, a montage loop of four output fields is displayed and updated regularly.
3. Comparison of Objectives vs Actual Accomplishments for Reporting Period:
In progress; the accomplishments for this project during this fiscal year compare favorably with the goals outlined in the statement of work.
4. Leveraging/Payoff:
5. Research Linkages/Partnerships/Collaborators:
6. Awards/Honors:
7. Outreach:
8. Publications:
Project Title: T-REX (Terrain-induced Rotor EXperiment)
Participating CIRA Researcher: Brian Jamison
NOAA Project Goal/Program: Commerce and Transportation—Support the Nation’s commerce with information for safe, efficient, and environmentally sound transportation/ Aviation weather
Key Words: T-REX, Terrain-induced Rotor EXperiment, Wind Flow in Complex Terrain
1. Long-term Research Objectives and Specific Plans to Achieve Them:
The T-REX project is a testbed for high-resolution, nonhydrostatic weather models in order to provide accurate guidance in complex environments. The test area for T-REX is in the vicinity of the Sierra Nevada mountain range, the location of some of the highest terrain in the coterminus U.S. Tasks for this project include: display of several model variables on a number of isobaric and height levels, display of model and diagnostic parameters through defined cross sections, and development of interactive webpages to facilitate analysis.
2. Research Accomplishments/Highlights:

Scripts were written to display the model and diagnostic variables including: potential temperature, relative humidity, dewpoint, wind, and vertical velocity for five pressure levels; wind, omega, potential temperature, and relative humidity for six height levels; and wind, omega, turbulent kinetic energy (TKE), Richardson number, Scorer parameter, Na/U and Nh/U (which describe the atmospheric response to orographic gravity waves) for vertical cross sections. Some additional plots are also generated for accumulated precipitation, mean sea level pressure, terrain, and skew-T charts for two locations. These scripts were implemented into a master script to automatically generate these graphic products following runs of two separate Weather Research and Forecasting (WRF) models: Advanced Research WRF (ARW) and National Mesoscale Model (NMM). Both of these models used a 2 km X 2 km domain with 50 vertical levels. The products are made available for viewing and comparison on a webpage (http://www-frd.fsl.noaa.gov/mab/trex/).

Adjustments were made to the aforementioned webpage and scripts. The webpage was modified to accept 27 different runs of the two WRF based models. This was designed to allow easy comparison of the output of the models following particular model parameter adjustments. Some other plots were added, including actual terrain along the cross section paths and four-panel plots which feature a top-down view of each relevant flight segment, the "crosstrack" plot (i.e. an aircraft track plotted on a vertical cross section), the comparison plot of aircraft observed theta and vertical velocity against the model data, and a model terrain plot along the aircraft path (Fig. 1).


Fig. 1. An example of a 4-panel plot, showing the plan view aircraft track, the "crosstrack" plot of the aircraft track on a vertical cross section, a comparison plot of aircraft and WRF-ARW vertical velocity, and plots of model and actual terrain.

Jim Doyle of the Naval Research Laboratory organized a T-REX model intercomparison study using 2-dimensional model runs with common initial conditions. From the WRF-ARW output, several products were generated including cross section plots of potential temperature, wind components, vertical velocity, turbulent kinetic energy, and eddy diffusivity. In all, five different simulations were performed, three using a simulated "Witch of Agnesi" mountain (Fig. 2), and the other two using actual terrain. Two of the simulation runs revealed some strange results, particularly in the plots of TKE. Further investigation revealed a "bug" in the WRF-ARW code, that when corrected, produced much more reasonable results. Using the corrected version of the model, the plots for all the simulations were redone. An internal webpage was created to allow easier analysis of both the "new" and "old" plots, which currently resides at http://www-frd.fsl.noaa.gov/mab/trex/2dsims.


Fig. 2. A 4-hour forecast of the east-west wind component for a 2D WRF-ARW simulation using a "Witch of Agnesi" mountain with a height of 1 kilometer. The color scale is wind magnitude in m/s and black contours are potential temperature.
3. Comparison of Objectives vs Actual Accomplishments for the Report Period:
In progress; the achievements for this project during this fiscal year compare favorably with the goals outlined in the statement of work.
4. Leveraging/Payoff:
5. Research Linkages/Partnerships/Collaborators:
6. Awards/Honors:
7. Outreach:
8. Publications:

II.


II. Research Collaborations with the GSD Aviation Branch
A. High Performance Computing-Advanced Computing
Project Title: Advanced High-Performance Computing
Principal Researchers:Tom Henderson, Jacques Middlecoff, Jeff Smith, and Ning Wang
NOAA Project Goals/Programs:
In the area of High Performance Computing-Advanced Computing, CIRA proposed seven research efforts. All seven efforts support NOAA mission goals of (1) Weather and Water—Serve society's needs for weather and water information / Environmental modeling; (2) Commerce and Transportation—Support the Nation's commerce with information for safe, efficient, and environmentally sound transportation / Aviation weather, Surface weather, and NOAA emergency response; and (3) Understand Climate Variability and Change to Enhance Society's Ability to Plan and Respond / Climate predictions and projections.
Key Words: Computational Grid, Gridpoint Statistical Interpolation, Model

Parallelization, WRF portal, FIM

1. Long-term Research Objectives and Specific Plans to Achieve Them:

a) CIRA researchers will continue to support NOAA code interoperability, as

well as ESRL needs, by supporting the NCEP Gridpoint Statistical Interpolation (GSI) code. The GSI is constantly changing, with a new version being released approximately every two months. CIRA researchers will insure that the latest version of GSI is available and ported to ijet, ejet and wjet. CIRA researchers will assist in coding changes necessary to enhance and optimize the GSI to prepare it for regional- and global-scale modeling activities.


b) CIRA researchers will also continue to develop and improve the WRF Portal—the graphical front end to the WRF NMM and ARW model. Specifically, they will generalize WRF Portal’s work flow management to support a wider variety of WRF tasks including the WRF Pre-processor System (WPS), and support additional HPC systems including ejet and wjet. They will also assist in expanding the use of WRF Portal to support other models including FIM, GFS, and GSI.
c) CIRA researchers will continue developing the computer science aspects of the Flow-following, Finite-volume Icosahedral Model (FIM). As FIM becomes more mature, emphasis will shift from development to optimization (both scalar and parallel) and testing.
d) CIRA researchers will continue their collaborations on the development of computer software for the parallelization of atmospheric and oceanic weather and climate models. Collectively, this software suite is known as the Scalable Modeling System (SMS). CIRA researchers will collaborate on adding SMS debugging capabilities to the WRF code.

e) CIRA researchers will also complete the development of Domain Wizard, ­a Java-based tool that replaces the existing WRFSI GUI (written in Perl). Domain Wizard enables users to graphically select and localize a domain for WRF ARW and NMM. CIRA researchers will continue to collaborate with developers of WPS at NCAR to ensure that the tool will be compatible with the newest version of WRF.

f) CIRA researchers will collaborate with ESRL scientists to support their codes and be available to provide design advice and expertise on a variety of software / web / database technologies for incorporation into the Lab’s various research endeavors. They will also assist in the research and development of web services that are expected to be used to locate and retrieve model data within GSD.
g) Finally, CIRA researchers will continue teaching Java courses to staff at ESRL as needed to support project development activities within the laboratory.
2. Research Accomplishments/Highlights:
Objective A:
The GSI code uses MPI2 I/O to enable multiple processors to write simultaneously to a single file. CIRA researchers investigated how this use of MPI2 I/O impacts the wjet file system and concluded that the wjet file system cannot support this I/O method when more than about 50 processors are used. Consequently, the wjet system managers came up with RAM disk as a method to support MPI2 I/O. So far, RAM disk has shown great promise for running GSI on large numbers of processors. CIRA researchers continue to fix bugs in GSI and help meteorologists who encountered problems with GSI.
Objective B:
In April 2008, CIRA researchers released a beta version of a Java application called WRF Portal that is a graphical user interface (GUI) to WRF-NMM and WRF-ARW and runs on most computer systems. WRF Portal also supports 2D visualization and it includes WRF Domain Wizard as a built-in software component. CIRA researchers created the wrfportal.org website, presented a paper at AMS, co-authored another paper at AMS, and gave presentations at the 2008 Winter WRF-NMM tutorial and plan on presenting a paper at the 9th Annual WRF Users Workshop in June 2008.

Objective C:

CIRA researchers have been an integral part of the team that has brought the FIM code to the current level of maturity where FIM produces daily weather forecasts. Specifically, CIRA researchers have created a complete system where one make command builds the pre-processing software, the FIM model, and the post processing software, and one qsub command runs the pre-processing, FIM and post-processing. CIRA researchers continue to improve and optimize FIM. Specifically, CIRA researchers are investigating the cache coherence of the unstructured grid and the advantages and disadvantages of traversing the globe using the Hilbert space-filling curve versus traversing the globe using Cartesian coordinates. In addition, CIRA researchers have improved the software engineering processes used during FIM development by creating source code repositories, developing an automated test suite for FIM, and implementing a lightweight software engineering process tailored to FIM requirements. The repositories are hosted by GSD’s GForge-based Subversion server. CIRA researchers taught an introductory course covering fundamentals of software revision control using Subversion and initially assisted FIM scientists and developers in its day-to-day use. With the new process, test suite, and repositories in place, FIM software engineering practices are now on a par with other major production NWP and climate codes such as CCSM and WRF. CIRA researchers refactored upper levels of FIM software to allow interoperability with NCEP’s NEMS architecture implemented via the ESMF. This effort included detailed collaboration with Tom Black at NCEP to generalize the NEMS ESMF approach so it meets requirements of NCEP models (GFS, NMMB) as well as FIM. CIRA researchers have created a prototype FIM ESMF component and are implementing the functional details to permit coupling of FIM physics and dynamics within the NEMS architecture. CIRA researchers also assisted GSD scientists to incorporate aspects of WRF-CHEM into FIM. This new feature now works correctly for serial execution, and distributed-memory parallel capability is under development.

In the grid generation package, CIRA researchers continue their work to improve the efficiency and flexibility of the software. The program that generates the grid mesh and cell boundaries is improved so that it will be able to accommodate different numerical schemes. The utilities for the package now include more tools to help trace and debug computations on the grid.
The post-processing package has been running in real-time for more than 3 months. Its output has been used to create display products for Fx-net, ALPS, and the SOS system. In March, these displays have been used to successfully demonstrate the FIM model to a high level delegation from NOAA headquarters.
Objective D:
CIRA researchers continue to improve SMS and to assist SMS users with SMS. SMS support for unstructured grids, newly developed last year by CIRA researchers in support of the FIM model, continues to be enhanced and optimized.
Objective E:
On April 17, 2008, CIRA researchers released version 1.10 of WRF Domain Wizard—the GUI for the new WRF Preprocessing System (WPS) and a component of WRF Portal. The new version supports editing namelist.input: the namelist file used to configure running real.exe and wrf.exe. WRF Domain Wizard enables users to choose a region of the Earth to be their domain, re-project that region in a variety of map projections, create nests using the nest editor, and run the three WPS programs. CIRA researchers are exploring the possibility of adapting this software to work with LEAD (Linked Environments for Atmospheric Discovery), a NSF-funded project.
Objective F:

CIRA researchers continue to help users parallelize and debug their codes on the jet computer systems, help ITS debug wjet, and improve the FIM part of the jet benchmarking suite. CIRA researchers have developed an OGC-compliant Data Locator web service (and website) for searching and viewing meteorological datasets here at GSD/ESRL. Utilizing a MySQL database and written primarily in Java, this website won a web award for 2007. CIRA researchers have also attended preliminary meetings on the probabilistic forecasting software project.
Objective G:
No Java classes have been taught since the summer of 2007.
3. Comparison of Objectives vs Actual Accomplishments for the Report Period:



  1   2   3   4   5   6   7   8   9   ...   13


The database is protected by copyright ©hestories.info 2017
send message

    Main page