MODIS Science Data Support Team Presentation Ed Masuoka Code 920.2 GSFC/NASA el~~astloka @ltpst]n. gsfc.llasa. gov AND Dr. Albert Fleig Department of Geography University of’ Maryland College Park :\fleig@ltpslln. gsfc. nasa. gov MODIS Science Team Meeting 10/12 /9’1
31
Embed
Science Data Support Team PresentationPresentation Ed Masuoka Code 920.2 GSFC/NASA el~~astloka @ltpst]n. gsfc.llasa. gov AND Dr. Albert Fleig Department of Geography University of’
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
MODISScience Data Support Team
Presentation
Ed MasuokaCode 920.2GSFC/NASA
el~~astloka @ltpst]n. gsfc.llasa. gov
AND
Dr. Albert FleigDepartment of Geography
University of’ Maryland College Park:\fleig@ltpslln. gsfc. nasa. gov
MODIS Science Team Meeting 10/12 /9’1
●
●
●
●
●
SDST
10/94
Topics
Contractor Support
Modeling Exercise
Ancillary Data Set Needs
Q/A Plans
12/94 SIW Readiness ReviewMODIS Science Team Meeting 10/12/94
HAIS Modeling Effort 10/94
● MODIS input to model has TBD’s for someparameters. These will end up zeroed outfor processing or storage.
● Modeling supports:analysis of product growth – Mel Bankspresentation to IWG - Bruce Barkstromcosting of ECS for PDR – HAIS/ECS
● MODIS July 1994 was 3.6 GFLOPS and is now4.9 GFLOPS. The 1.6x multiplier we had inprevious updates was dropped from thelast SPSO database update. Storagevolume grew from 938 GB to 1,082 GB.
MODIS Science Team Meeting 10/12/94
__. — —-— ...... .—
Parameter
Ancillary DataCIOLICi Product
Description
Global Surface Parameters(Temp, Pres, and wind speecI)
Reynolds Blendecl SST— ——.—— ————-.—-
Global Model Profiles(Temp and Moisture)
—
Vegetation Type
Digital Elevation Map
AI RS/AMSU Profiles(Temp and Moisture)
TangibleSource’
1--”--——.—..
NMC
NMC———.
NMC
—..._._——-
GSFCIDAAC
—.—— —-.—.—.—
‘ft~ server and Dathname, archive site, CD ROM
—.—.——-—- .-—-—~.-—.——. ..–——
Type , A~plic~ble Epoch(P/V/D)2 ; (BL/AL/PL)3
p
P
P
P
P——
P
— —.————
ALIPL
AL/PL—.—..—-.—————.—
ALIPL
AL/PL———-——.. ——
ALIPL
PL
2P;oduction (P),’ validation (V), algorithm cievelopment (D)3Before Launch (BL), At Launch (AL), and Post-Launch (PL)
QIA Plans
● Robert Lutz common Q/A for AM instrumentscomments from Evans and Masuoka
● Need to work Q/A plan for MODIS productsImpact on wide area networksQ/A activities at DAAC and SCFCommonality with other instruments
● Needed for data management plan revision inearly November
MODIS Science Team Meeting 10/12/94
A software readiness review ofMODIS will take place in December1994 for Dixon Butler and willcover:
Products to be generated, their phasing, and readiness plans
Software implementation plan and status
Resource requirements for development (hardware and personnel)and status
External interfaces and status (e.g., data dependencies, tool kits,network requirements)
Test plans
QA plans, procedures and resource implications
Plans for receiving and integrating Team Members’ software at theTeam Leader Computing Facility
Plans for participation in integration of MODIS software into theEOSDIS product generation system
Resource requirements for product generation on EOSDIS
Significant problems/issues
This is intended to be more a status re~iew than an algorithm,science or software design review. Sufficient detail should beprovided commensurate u)ith the time to demonstrate that allrelevant developmental issues have been identified and are beingaddressed.
THANKING YOU IN ADVANCE FORYOUR SUPPORT.
Topics
● Flathead Lake - Al Fleig
. Validation Plan - Al Fleig
MODIS Science Team Meeting 10/12/94
Topics
● Beta Software Delivery
● SDST Utilities
● Nested ISSCP Grids
Q LIB Rectified/Mapped
● DAAC Activities
MODIS Science Team Mee[ing 10/12/94
Beta Software Deliveries
● 1/94 Heritage SIWWork the del iverv processDeve 10P ~lans for heritaae s/w
The MODIS scan cube data structure canbe complex to handle due to differencesin spatial resolution between bands and acommendable number of Science andCalibration frames.
I I Library handles checking thatLibrary sets up output data MOD 06 Cloud Product correct data set is read andsets in structures specified Science Algorithm
by include file to simplifysets up pointers to parametersso access can be by name
input by other algorithms
1{, Nlilsll(lLik/sl )s’r/l~l;llll(’iltl I .clA[/’)/’)l
To Use the Utility Library You Need:
●
●
●
●
Include files for each data set that will bewritten by the library. SDST can help.
The names of variables in each data set
Information about each variable (units, howcalculated)
The utilities are written in C (SDST will need todevelop a wrapper if you are deliveringFORTRAN 77 or FORTRAN 90)
MODIS Science Tean~ Meeting 10/94
But there is also HDF.
● Self describing data structures
●Approved by ECS for archival
and production
● Supports viewers
● May support our scan cubeMODIS Science Teanl Meeting !O/ 12/94
I
}
Data Type Taxonomy
TSeienw
DalaTable
II
I
1~ II
“< Access
Does Code Image Tables
1OIR, ItiV
I Oti II 1 Wm
r 1 I 11 FlodiKth
ASCIIText
tlktqDintary ASCII Bintafy 8-bit 24-bil ~ SUHH
I Gti ALG U ~, ~
d ~ stats
Swath
● Simple
- Created by a sensor making N observations in theacross-track direction. The along-track directioncauses the footprint to form a ribbon of M scans aIongthe subnadir track. The data forms an array ofobservations N by M by L (where L is the number ofdata channels taken for each observation time). Anadditional array of gee-location or observation timedata is provided at a resolution equal to or lower thanthe observations
● Indexed
– Similar to the Simple Swatl~ but may haveobservations of varying resolution
Sensor
cl+
●
GridStructured ● Rectangular
- Data which has been projected and binnedinto a non-rectilinear data structure using aknown methodology. Metadata such asprojection name, projection limits, &geometry are included in order to identifygee-location and coverage of data structurecells
– Data which has been projected andbinned into a rectangular grid ~]sing aknown methodology. Metadata such asprojection name, projection limits, &geometry are included in order toidentify geo-Iocation and coverage ofgrid ceils.
ISSCP Grid - Nested Cells (from 9.28 km down)—
18.55 km(1Oarc rein)
L’
I
I
0.93 km ~ –
1.28 km; arc rein)
*
:30 arc see) ~
0.46 km \ 4.64 km(15 arc see) (150 arc see)
__— ——.———.—
_—.-———
—
10/11 /94 Robert Wolfe/MOfllS SDS r
ISSCP Grid Nesting Issues
Nesting Advantages
Only Needed Horizontally - ISSCP Grid Naturally Nested Vertically
Boundaries Line up Nicely Within 9.28 km Area - Interpolation Easier
Area Based Compression (Quad tree) Can be Used
Is 9.28 km the “best” nesting size?
Satisfies Equal Area Criteria
Oceans Currently Use 9.28 km Grid Size
Most Land/Ocean Grid Sizes Handled - Land/Sea Boundary
Atmosphere may Use 18.55 km or Larger Grid
Indexing Uses Hybrid Approach
Simplifies Indexing within Nested Area
Uses Standard ISSCP Indexing when Crossing Upper/Lower Boundaries
Roberi Wolfe/MODIS SDST
MODIS Scanning Geomet~
/
MODIS/ScienceTeam Meeting Page -2 V. L. Wb, T. E. Goff