Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS)

Alison O'Connor, Benjamin Kirtman, Scott Harrison, Joe Gorman

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The US Navy faces several limitations when planning operations in regard to forecasting environmental conditions. Currently, mission analysis and planning tools rely heavily on short-term (less than a week) forecasts or long-term statistical climate products. However, newly available data in the form of weather forecast ensembles provides dynamical and statistical extended-range predictions that can produce more accurate predictions if ensemble members can be combined correctly. Charles River Analytics is designing the Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS), which performs data fusion over extended-range multi-model ensembles, such as the North American Multi-Model Ensemble (NMME), to produce a unified forecast for several weeks to several seasons in the future. We evaluated thirty years of forecasts using machine learning to select predictions for an all-encompassing and superior forecast that can be used to inform the Navy's decision planning process.

Original languageEnglish (US)
Title of host publicationModeling and Simulation for Defense Systems and Applications XI
PublisherSPIE
Volume9848
ISBN (Electronic)9781510600898
DOIs
StatePublished - 2016
EventModeling and Simulation for Defense Systems and Applications XI - Baltimore, United States
Duration: Apr 18 2016 → …

Other

OtherModeling and Simulation for Defense Systems and Applications XI
CountryUnited States
CityBaltimore
Period4/18/16 → …

Fingerprint

forecasting
planning
Forecast
Planning
Ensemble
Prediction
predictions
Multi-model
navy
Process Planning
Data Fusion
Data fusion
machine learning
Climate
Weather
multisensor fusion
Range of data
Learning systems
Forecasting
Machine Learning

Keywords

  • climate predictability
  • Extended-range forecasting
  • Figaro programming language1
  • machine learning
  • multi-model ensembles
  • probabilistic forecasting
  • probabilistic programming

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Condensed Matter Physics
  • Computer Science Applications
  • Electrical and Electronic Engineering
  • Applied Mathematics

Cite this

O'Connor, A., Kirtman, B., Harrison, S., & Gorman, J. (2016). Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS). In Modeling and Simulation for Defense Systems and Applications XI (Vol. 9848). [98480D] SPIE. https://doi.org/10.1117/12.2223328

Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS). / O'Connor, Alison; Kirtman, Benjamin; Harrison, Scott; Gorman, Joe.

Modeling and Simulation for Defense Systems and Applications XI. Vol. 9848 SPIE, 2016. 98480D.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

O'Connor, A, Kirtman, B, Harrison, S & Gorman, J 2016, Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS). in Modeling and Simulation for Defense Systems and Applications XI. vol. 9848, 98480D, SPIE, Modeling and Simulation for Defense Systems and Applications XI, Baltimore, United States, 4/18/16. https://doi.org/10.1117/12.2223328
O'Connor A, Kirtman B, Harrison S, Gorman J. Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS). In Modeling and Simulation for Defense Systems and Applications XI. Vol. 9848. SPIE. 2016. 98480D https://doi.org/10.1117/12.2223328
O'Connor, Alison ; Kirtman, Benjamin ; Harrison, Scott ; Gorman, Joe. / Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS). Modeling and Simulation for Defense Systems and Applications XI. Vol. 9848 SPIE, 2016.
@inproceedings{ae11eb74265548c0bcc1bafc115c72ae,
title = "Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS)",
abstract = "The US Navy faces several limitations when planning operations in regard to forecasting environmental conditions. Currently, mission analysis and planning tools rely heavily on short-term (less than a week) forecasts or long-term statistical climate products. However, newly available data in the form of weather forecast ensembles provides dynamical and statistical extended-range predictions that can produce more accurate predictions if ensemble members can be combined correctly. Charles River Analytics is designing the Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS), which performs data fusion over extended-range multi-model ensembles, such as the North American Multi-Model Ensemble (NMME), to produce a unified forecast for several weeks to several seasons in the future. We evaluated thirty years of forecasts using machine learning to select predictions for an all-encompassing and superior forecast that can be used to inform the Navy's decision planning process.",
keywords = "climate predictability, Extended-range forecasting, Figaro programming language1, machine learning, multi-model ensembles, probabilistic forecasting, probabilistic programming",
author = "Alison O'Connor and Benjamin Kirtman and Scott Harrison and Joe Gorman",
year = "2016",
doi = "10.1117/12.2223328",
language = "English (US)",
volume = "9848",
booktitle = "Modeling and Simulation for Defense Systems and Applications XI",
publisher = "SPIE",

}

TY - GEN

T1 - Operational planning using Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS)

AU - O'Connor, Alison

AU - Kirtman, Benjamin

AU - Harrison, Scott

AU - Gorman, Joe

PY - 2016

Y1 - 2016

N2 - The US Navy faces several limitations when planning operations in regard to forecasting environmental conditions. Currently, mission analysis and planning tools rely heavily on short-term (less than a week) forecasts or long-term statistical climate products. However, newly available data in the form of weather forecast ensembles provides dynamical and statistical extended-range predictions that can produce more accurate predictions if ensemble members can be combined correctly. Charles River Analytics is designing the Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS), which performs data fusion over extended-range multi-model ensembles, such as the North American Multi-Model Ensemble (NMME), to produce a unified forecast for several weeks to several seasons in the future. We evaluated thirty years of forecasts using machine learning to select predictions for an all-encompassing and superior forecast that can be used to inform the Navy's decision planning process.

AB - The US Navy faces several limitations when planning operations in regard to forecasting environmental conditions. Currently, mission analysis and planning tools rely heavily on short-term (less than a week) forecasts or long-term statistical climate products. However, newly available data in the form of weather forecast ensembles provides dynamical and statistical extended-range predictions that can produce more accurate predictions if ensemble members can be combined correctly. Charles River Analytics is designing the Climatological Observations for Maritime Prediction and Analysis Support Service (COMPASS), which performs data fusion over extended-range multi-model ensembles, such as the North American Multi-Model Ensemble (NMME), to produce a unified forecast for several weeks to several seasons in the future. We evaluated thirty years of forecasts using machine learning to select predictions for an all-encompassing and superior forecast that can be used to inform the Navy's decision planning process.

KW - climate predictability

KW - Extended-range forecasting

KW - Figaro programming language1

KW - machine learning

KW - multi-model ensembles

KW - probabilistic forecasting

KW - probabilistic programming

UR - http://www.scopus.com/inward/record.url?scp=84982283218&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84982283218&partnerID=8YFLogxK

U2 - 10.1117/12.2223328

DO - 10.1117/12.2223328

M3 - Conference contribution

VL - 9848

BT - Modeling and Simulation for Defense Systems and Applications XI

PB - SPIE

ER -