Features
Get A Demo
Plans
Sign Up
Login
All Federal Grant Awards
Help
Project Grant NA21NMF4720289
Award Date
10/1/21
Completion Date
9/30/26
Dollars Obligated
$19M
Overview
Activity
Transactions
Subawards
61
Federal Agency
National Oceanic and Atmospheric Administration
Awardee
North Pacific Research Board (ENNLMJGGG9L1)
Federal Grant Program
11.472
Assistance Type
Project Grant
Place of Performance
Anchorage, AK 99501, USA
Description
Update #1
NORTH PACIFIC RESEARCH BOARD, 2021-2026
Posted 6/14/21, 12:00 AM
Grant Number
Description
Subgrantee
Prime Award
Dollars Obligated
(Click to sort descending)
Updated At
(Click to sort ascending)
2210AF218002S
"ALASKA SALMON STOCKS HAVE RESPONDED DIFFERENTLY TO RECENT CHANGES IN THE NORTH PACIFIC OCEAN (NPO). RECORD LOW RUN SIZES OF CHUM SALMON WERE OBSERVED IN THE YUKON RIVER AND NORTON SOUND DURING 2021, BUT RECORD HIGH RUN SIZES OF SOCKEYE SALMON WERE PRESENT IN BRISTOL BAY. THE DIFFERENTIAL SURVIVAL RESPONSE OF SALMON TO CHANGES IN THE MARINE ECOSYSTEM HIGHLIGHTS THE IMPORTANCE OF UNDERSTANDING SPECIES AND STOCK LEVEL DIFFERENCES IN THE MARINE ECOLOGY OF SALMON AND ITS IMPACT ON CRITICAL SURVIVAL PERIODS. WE PROPOSE A FOCUSED STUDY ON THE MARINE ECOLOGY OF YUKON RIVER AND NORTON SOUND CHUM SALMON BY EXAMINING CRITICAL PERIODS DURING THEIR FIRST YEAR AT SEA AND OVER WINTER. EARLY MARINE LIFE HISTORY DATA COME FROM A TIME SERIES (2002 TO PRESENT) OF BIO/PHYSICAL OCEANOGRAPHIC AND FISH COLLECTIONS ON THE NORTHERN BERING SEA SHELF. WINTER MARINE ECOLOGY DATA COME FROM INTERNATIONAL YEAR OF THE SALMON (IYS) SURVEYS IN THE GULF OF ALASKA (GOA) DURING 2019 AND 2020. AN IYS SURVEY WITHIN THE NPO BY PARTIES TO THE NORTH PACIFIC ANADROMOUS FISH COMMISSION DURING WINTER 2022 WILL ENABLE COLLECTIONS OF BIO/PHYSICAL OCEANOGRAPHIC DATA AND PACIFIC SALMON ON A BROADER SCALE. THESE DATA WILL BE USED TO EXAMINE CLIMATE IMPACTS ON CRITICAL PERIODS FOR WESTERN ALASKA CHUM SALMON DURING THEIR EARLY LIFE HISTORY STAGE TO THEIR FIRST WINTER IN THE NPO BY TESTING THE FOLLOWING HYPOTHESES: 1) THE GOA AND EASTERN NPO SERVE AS WINTER HABITAT FOR IMMATURE WESTERN ALASKA CHUM SALMON STOCKS; 2) WINTER CONDITIONS AND FORAGE RESOURCES FOR IMMATURE CHUM SALMON ARE SUFFICIENT TO STAVE OFF STARVATION.; AND 3) COMPETITION AMONG SALMON IS GREATER DURING WINTER RELATIVE TO OTHER COMPETITIVE PRESSURES."
University Of Washington
Project Grant NA21NMF4720289
$42.1k
2/8/24
H3281S
ECOHAB Student
Aleut Community Of Saint Paul Island
Project Grant NA21NMF4720289
$70.8k
8/23/22
F2180032210S
YUKON RIVER AND NORTON SOUND CHUM SALMON MARINE ECOLOGY
University Of Alaska Fairbanks
Project Grant NA21NMF4720289
$33.1k
1/25/23
F2128022208S
ALASKAN PINNIPED DIVING CONSTRAINTS AND ADAPTIVE CAPACITY
Alaska Marine Science Association, LLC
Project Grant NA21NMF4720289
$47.4k
1/19/23
2410F214000S
"BETWEEN 2007 AND 2014 THE SHELL EXPLORATION AND PRODUCTION COMPANY FUNDED GREENERIDGE SCIENCES, A BIOACOUSTICS CONSULTING FIRM, TO DEPLOY 35 PASSIVE ACOUSTIC DIRECTIONAL RECORDERS ALONG A 280 KM SWATH OF THE ALASKAN NORTH SLOPE DURING THE ANNUAL BOWHEAD WHALE FALL MIGRATION. THE SENSORS DETECTED AND LOCALIZED BOWHEAD WHALE CALLS OF VARIOUS TYPES. EACH SEASON A TEAM OF MANUAL ANALYSTS ANNOTATED THE TIME, FREQUENCY RANGE, DISTANCE, AND TYPE OF BOWHEAD CALLS, EVENTUALLY LABELING 8.7 MILLION MANUALLY ANNOTATED CALLS, ONE OF THE LARGEST ANNOTATED BOWHEAD WHALE SOUND DATASETS IN EXISTENCE. IN 2010 A SIMPLE NEURAL NETWORK FOR IDENTIFYING BOWHEAD CALLS WAS DEVELOPED AND APPLIED TO LOCALIZE MILLIONS OF BOWHEAD WHALE CALLS ACROSS THE ENTIRE PROJECT. HOWEVER, THE THREE-LAYER Â SHALLOW LEARNINGÂ NETWORK COULD ONLY RECOGNIZE, AND NOT CLASSIFY, RELATIVELY SIMPLE FREQUENCY-MODULATED SOUNDS, AND NOT MORE COMPLEX SOUNDS PRODUCED BY THE ANIMALS. ADVANCES IN MACHINE LEARNING, PARTICULARLY "DEEP LEARNING" CONVOLUTIONAL NEURAL NETWORKS, NOW MAKE POSSIBLE MUCH MORE ADVANCED DETECTION AND CLASSIFICATION OF BOWHEAD WHALE CALLS. THIS PROJECT WILL USE THIS LARGE SUPERVISED MANUAL DATA SET TO APPLY TRANSFER LEARNING TO AN EXISTING DEEP LEARNING CONVOLUTIONAL NETWORK ARCHITECTURE TO CREATE A MODERN MACHINE LEARNING APPLICATION THAT WILL DETECT AND CLASSIFY BOWHEAD WHALE CALLS INTO UP CALLS, DOWN CALLS, UNDULATIONS, TONAL, AND COMPLEX CALLS, AS WELL AS ESTIMATE A CALLÂ S RANGE. THE ALGORITHM WILL BE TESTED ON MANUALLYANNOTATED DATASETS COLLECTED BY GREENERIDGE, OTHER SHELL-SPONSORED ACOUSTIC DEPLOYMENTS IN THE CHUKCHI SEA, AND OTHER PASSIVE ACOUSTIC DATA SETS COLLECTED BY THE NOAA MARINE MAMMAL LABORATORY. THE EXISTENCE OF A DEDICATED BOWHEAD WHALE SOUND CLASSIFIER, WHEN MADE AVAILABLE TO ANALYZE HUGE EXISTING PASSIVE ACOUSTIC DATASETS, WILL PROVIDE NEW INSIGHTS INTO THE FUNCTION OF VARIOUS CALL TYPES AND LONG-TERM CHANGES IN GEOGRAPHIC DISTRIBUTION AND CALL REPERTOIRE IN RESPONSE TO CLIMATE CHANGE AND HUMAN DISTURBANCE."
University Of California San Diego
Project Grant NA21NMF4720289
$150.0k
12/2/24