SBIR-STTR Award

The objective of the proposed effort is to investigate the feasibility of an end-to-end global long-term ionospheric forecast model based on a fusion of several diverse technologies and to develop the related probability density function evolution formalism to characterize the forecast quality. In order to meet the stated goal of a 3-day forecast one has to address the complete chain of events starting from highly unpredictable changes in solar conditions to changes in the ionosphere. Ideally, the system would consist of several physics-based models, a sufficient number of observational data streams and a data assimilation system that provides for computing error covariance evolution. Presently, an end-to-end first-principles based assimilative system is impossible. We propose a practical system based on a synthesis of several different technologies: (1) an artificial intelligence algorithm known as Support Vector Machines for predicting changes in solar wind from time sequences of solar images; (2) an empirical model of the high-latitude electric field potentials; and (3) a physics-based ionospheric model coupled with efficient Kalman filter for forecasting the final ionospheric parameters of interest. Additionally, we propose a prototype error propagation scheme for computing evolution of forecast probability density functions starting from errors of representativeness in the synoptic solar images to uncertainties in the final forecast.

Benefits:
Improvements in space weather modeling and forecasting will be of immediate use for a number of practical military and civilian applications, particularly in satellite-based communications and navigation. Our commercialization strategy is based on the fact that contemporary space weather models are not capable of generating precise forecasts for use by those industries where solar and Ionospheric affects disrupt operations in a costly manner. At the same time, given our reliance and dependency on satellite and wireless communications such forecasts are of considerable interest to the private sector and the military to allow for operational planning instead of emergency reaction. In the private sector potential clients include: companies in satellite-based navigation (GPS industry); satellite-based communications, including high band width requirements and mission critical applications; cellular communications companies; power distribution concerns; and research institutions. Development of a physics-based ionospheric forecast system will address these needs and open up radically new commercial and military applications. To further substantiate commercial application of this technology we have established relationships in the commercial sector with major GPS service companies, confirmed by the enclosed letters of interest.

Keywords:
ionosphere, forecasting, data assimilation, machine learning, space weather, ensemble filter

The main objective of this proposal is to provide a modeling, assimilation, forecasting, and data analysis platform to support the Communication / Navigation Outage Forecasting System (C/NOFS) mission of the AFRL. The system will combine several components developed by Fusion Numerics Inc. along with new specialized modules. The goal of the C/NOFS mission is to understand the physics of the ionosphere in equatorial regions and to forecast accurately the subsequent scintillation-producing irregularities. To meet this goal, we propose a system based on the synthesis of several unique techniques developed during prior SBIR funded projects: (1) an artificial intelligence algorithm known as Support Vector Machine for predicting changes in solar wind from solar images; (2) an empirical model of electric field potentials; (3) a physics-based assimilative ionospheric model for forecasting ionospheric dynamics; and (4) a model and forecast uncertainty evolution scheme based on ensemble Kalman filter. Additionally, we propose a novel infrastructure for a modular unified forecast system. Such an infrastructure will allow practitioners to easily replace models of ionospheric components as they become available. The infrastructure will rely on ESML (Earth Science Markup Language) and SOAP (Simple Object Access Protocol) based web-services for model interfacing, data stream integration and data distribution.

Benefits:
Our product has applications in most areas of military operations, from HF communications to delivering vital data to individual soldiers in remote battlefields, to precision guided weapons, to space based intelligence gathering. Many existing military operations will see immediate benefits, e.g., better position determination with GPS receivers and efficient HF communications. The unique ability to forecast regional ionospheric conditions will allow the military connectivity providers to predict potential communications interferences (situational awareness) and make proactive routing decisions or operate on different frequencies in response to the changing environment. While several DoD agencies will benefit from using the proposed technology, the US Air Force will likely see the biggest immediate cost-savings resulting from a better ability to forecast communication outages as well as quiet conditions suitable for critical mission initiation. Target commercial applications of our product include HF FAA communications, remote voice and data services, precise mapping, E-911 services, intelligent automated vehicles (terrestrial and airborne), goods and supplies distribution management, and modern agriculture.

Keywords:
ionosphere, scintillations, space weather, GPS, communications, navigation, forecast, mapping