Aurorasaurus: Citizen science, early warning systems and space weather

Andrea H. Tapia, Nicolas Lalone, Elizabeth MacDonald, Nathan Case, Michelle Hall, Matt Heavner

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

3 Citations (Scopus)

Abstract

We have created Aurorasaurus, a website, a mobile application and a scientific tool that allows a community of users to better predict sightings of the aurora borealis. We focus on the aurora borealis as a rare and unpredictable event (as a proxy for a natural disaster), as it is in the middle latitudes, highly populated areas in North American and Europe, in the northern half of the continental United States the aurora may be visible once or twice per year. In the Southern half, perhaps only once every 20 years, especially during a solar maximum. We feel that the similarities between natural disaster occurrence and auroral occurrence can offer a chance for researchers to test elements of an Early Warning System. The years around 2014 arc the latest solar maximum recurring on an eleven-year solar cycle. Now is the time when aurora is more likely over populated areas, and this is the first solar inaximum with social media, an unprecedented opportunity to engage the public, the scientific community, and the media. Space weather scientists have only coarse means to predict where the aurora will occur. Forecasts derived from state-of-the-solar wind models and satellite-based observations of the Sun estimate the arrival of the coronal mass ejection material at Earth to a typical accuracy of ±8 hours. There is no way to accurately predict the strength of an incoming storm. Space weather observations are simply not plentiful enough to accurately forecast the 93 million miles of interplanetary plasma between the Sun and the Earth. However, a network of observers on the ground can provide ground truth, which could enable nowcasts. We anticipate an order of magnitude increase in accurate knowledge of where and when the aurora are visible by augmenting existing observations and models with real-time ground truth (thus going from 100s to 10s of km, and from hour to minute time-scales). Using these new data we anticipate three major impacts on geospace science: 1) Significantly more accurate, timely, and relevant methods of auroral predictions. 2) New cost-effective sources of human-derived data for observing the aurora, including volunteered geographic information, images and videos, which may aid scientists in understanding their space-based data. 3) The ability to study auroras at mid-latitudes during storms. A storm has never been imaged at mid-latitudes and the last global imaging satellites with significantly lower resolution ended their missions in 2006. Space scientists are still tillng to connect many magnetosphcric regions to their visible manifestations in the ionosphere in this data-starved, vast, and relatively young field. Other citizen science activities may yield discoveries on the shape of the dynamic auroral oval versus time and activity due to the current lacic of data on these phenomena. On October 24th, 2011 aurora was visible as far south as Alabama. Due to the excelletit timing and widespread visibility in clear skies over heavily populated areas, this storm was covered by more than 500 traditional and online news outlets from large international newspapers to blogs. In real-time, tweets exceeded one per second during the event. Additionally, thousands of Twitter messages (tweets) containing text and images documented the visible aurora in real-time (MacDonald, 2012). During the testing of our prototype, we found that the Kp index, a key measure of auroral activity, correlates strongly with the number of aurora-related tweets. This suggests that volunteered Twitter data is a valid indicator of real-world events.

Original languageEnglish (US)
Title of host publicationCitizen + X
Subtitle of host publicationVolunteer-Based Crowdsourcing in Science, Public Health and Government - Papers from the 2014 HCOMP Workshop, Technical Report
PublisherAI Access Foundation
Pages30-32
Number of pages3
VolumeWS-14-20
ISBN (Electronic)9781577356905
StatePublished - Jan 1 2014
Event2nd AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2014 - Pittsburgh, United States
Duration: Nov 2 2014 → …

Other

Other2nd AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2014
CountryUnited States
CityPittsburgh
Period11/2/14 → …

Fingerprint

Alarm systems
Sun
Disasters
Earth (planet)
Satellites
Solar wind
Blogs
Ionosphere
Visibility
Websites
Plasmas
Imaging techniques
Testing
Costs

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Tapia, A. H., Lalone, N., MacDonald, E., Case, N., Hall, M., & Heavner, M. (2014). Aurorasaurus: Citizen science, early warning systems and space weather. In Citizen + X: Volunteer-Based Crowdsourcing in Science, Public Health and Government - Papers from the 2014 HCOMP Workshop, Technical Report (Vol. WS-14-20, pp. 30-32). AI Access Foundation.
Tapia, Andrea H. ; Lalone, Nicolas ; MacDonald, Elizabeth ; Case, Nathan ; Hall, Michelle ; Heavner, Matt. / Aurorasaurus : Citizen science, early warning systems and space weather. Citizen + X: Volunteer-Based Crowdsourcing in Science, Public Health and Government - Papers from the 2014 HCOMP Workshop, Technical Report. Vol. WS-14-20 AI Access Foundation, 2014. pp. 30-32
@inproceedings{f3dfa8c6416049719a9006cc97184f6b,
title = "Aurorasaurus: Citizen science, early warning systems and space weather",
abstract = "We have created Aurorasaurus, a website, a mobile application and a scientific tool that allows a community of users to better predict sightings of the aurora borealis. We focus on the aurora borealis as a rare and unpredictable event (as a proxy for a natural disaster), as it is in the middle latitudes, highly populated areas in North American and Europe, in the northern half of the continental United States the aurora may be visible once or twice per year. In the Southern half, perhaps only once every 20 years, especially during a solar maximum. We feel that the similarities between natural disaster occurrence and auroral occurrence can offer a chance for researchers to test elements of an Early Warning System. The years around 2014 arc the latest solar maximum recurring on an eleven-year solar cycle. Now is the time when aurora is more likely over populated areas, and this is the first solar inaximum with social media, an unprecedented opportunity to engage the public, the scientific community, and the media. Space weather scientists have only coarse means to predict where the aurora will occur. Forecasts derived from state-of-the-solar wind models and satellite-based observations of the Sun estimate the arrival of the coronal mass ejection material at Earth to a typical accuracy of ±8 hours. There is no way to accurately predict the strength of an incoming storm. Space weather observations are simply not plentiful enough to accurately forecast the 93 million miles of interplanetary plasma between the Sun and the Earth. However, a network of observers on the ground can provide ground truth, which could enable nowcasts. We anticipate an order of magnitude increase in accurate knowledge of where and when the aurora are visible by augmenting existing observations and models with real-time ground truth (thus going from 100s to 10s of km, and from hour to minute time-scales). Using these new data we anticipate three major impacts on geospace science: 1) Significantly more accurate, timely, and relevant methods of auroral predictions. 2) New cost-effective sources of human-derived data for observing the aurora, including volunteered geographic information, images and videos, which may aid scientists in understanding their space-based data. 3) The ability to study auroras at mid-latitudes during storms. A storm has never been imaged at mid-latitudes and the last global imaging satellites with significantly lower resolution ended their missions in 2006. Space scientists are still tillng to connect many magnetosphcric regions to their visible manifestations in the ionosphere in this data-starved, vast, and relatively young field. Other citizen science activities may yield discoveries on the shape of the dynamic auroral oval versus time and activity due to the current lacic of data on these phenomena. On October 24th, 2011 aurora was visible as far south as Alabama. Due to the excelletit timing and widespread visibility in clear skies over heavily populated areas, this storm was covered by more than 500 traditional and online news outlets from large international newspapers to blogs. In real-time, tweets exceeded one per second during the event. Additionally, thousands of Twitter messages (tweets) containing text and images documented the visible aurora in real-time (MacDonald, 2012). During the testing of our prototype, we found that the Kp index, a key measure of auroral activity, correlates strongly with the number of aurora-related tweets. This suggests that volunteered Twitter data is a valid indicator of real-world events.",
author = "Tapia, {Andrea H.} and Nicolas Lalone and Elizabeth MacDonald and Nathan Case and Michelle Hall and Matt Heavner",
year = "2014",
month = "1",
day = "1",
language = "English (US)",
volume = "WS-14-20",
pages = "30--32",
booktitle = "Citizen + X",
publisher = "AI Access Foundation",
address = "United States",

}

Tapia, AH, Lalone, N, MacDonald, E, Case, N, Hall, M & Heavner, M 2014, Aurorasaurus: Citizen science, early warning systems and space weather. in Citizen + X: Volunteer-Based Crowdsourcing in Science, Public Health and Government - Papers from the 2014 HCOMP Workshop, Technical Report. vol. WS-14-20, AI Access Foundation, pp. 30-32, 2nd AAAI Conference on Human Computation and Crowdsourcing, HCOMP 2014, Pittsburgh, United States, 11/2/14.

Aurorasaurus : Citizen science, early warning systems and space weather. / Tapia, Andrea H.; Lalone, Nicolas; MacDonald, Elizabeth; Case, Nathan; Hall, Michelle; Heavner, Matt.

Citizen + X: Volunteer-Based Crowdsourcing in Science, Public Health and Government - Papers from the 2014 HCOMP Workshop, Technical Report. Vol. WS-14-20 AI Access Foundation, 2014. p. 30-32.

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

TY - GEN

T1 - Aurorasaurus

T2 - Citizen science, early warning systems and space weather

AU - Tapia, Andrea H.

AU - Lalone, Nicolas

AU - MacDonald, Elizabeth

AU - Case, Nathan

AU - Hall, Michelle

AU - Heavner, Matt

PY - 2014/1/1

Y1 - 2014/1/1

N2 - We have created Aurorasaurus, a website, a mobile application and a scientific tool that allows a community of users to better predict sightings of the aurora borealis. We focus on the aurora borealis as a rare and unpredictable event (as a proxy for a natural disaster), as it is in the middle latitudes, highly populated areas in North American and Europe, in the northern half of the continental United States the aurora may be visible once or twice per year. In the Southern half, perhaps only once every 20 years, especially during a solar maximum. We feel that the similarities between natural disaster occurrence and auroral occurrence can offer a chance for researchers to test elements of an Early Warning System. The years around 2014 arc the latest solar maximum recurring on an eleven-year solar cycle. Now is the time when aurora is more likely over populated areas, and this is the first solar inaximum with social media, an unprecedented opportunity to engage the public, the scientific community, and the media. Space weather scientists have only coarse means to predict where the aurora will occur. Forecasts derived from state-of-the-solar wind models and satellite-based observations of the Sun estimate the arrival of the coronal mass ejection material at Earth to a typical accuracy of ±8 hours. There is no way to accurately predict the strength of an incoming storm. Space weather observations are simply not plentiful enough to accurately forecast the 93 million miles of interplanetary plasma between the Sun and the Earth. However, a network of observers on the ground can provide ground truth, which could enable nowcasts. We anticipate an order of magnitude increase in accurate knowledge of where and when the aurora are visible by augmenting existing observations and models with real-time ground truth (thus going from 100s to 10s of km, and from hour to minute time-scales). Using these new data we anticipate three major impacts on geospace science: 1) Significantly more accurate, timely, and relevant methods of auroral predictions. 2) New cost-effective sources of human-derived data for observing the aurora, including volunteered geographic information, images and videos, which may aid scientists in understanding their space-based data. 3) The ability to study auroras at mid-latitudes during storms. A storm has never been imaged at mid-latitudes and the last global imaging satellites with significantly lower resolution ended their missions in 2006. Space scientists are still tillng to connect many magnetosphcric regions to their visible manifestations in the ionosphere in this data-starved, vast, and relatively young field. Other citizen science activities may yield discoveries on the shape of the dynamic auroral oval versus time and activity due to the current lacic of data on these phenomena. On October 24th, 2011 aurora was visible as far south as Alabama. Due to the excelletit timing and widespread visibility in clear skies over heavily populated areas, this storm was covered by more than 500 traditional and online news outlets from large international newspapers to blogs. In real-time, tweets exceeded one per second during the event. Additionally, thousands of Twitter messages (tweets) containing text and images documented the visible aurora in real-time (MacDonald, 2012). During the testing of our prototype, we found that the Kp index, a key measure of auroral activity, correlates strongly with the number of aurora-related tweets. This suggests that volunteered Twitter data is a valid indicator of real-world events.

AB - We have created Aurorasaurus, a website, a mobile application and a scientific tool that allows a community of users to better predict sightings of the aurora borealis. We focus on the aurora borealis as a rare and unpredictable event (as a proxy for a natural disaster), as it is in the middle latitudes, highly populated areas in North American and Europe, in the northern half of the continental United States the aurora may be visible once or twice per year. In the Southern half, perhaps only once every 20 years, especially during a solar maximum. We feel that the similarities between natural disaster occurrence and auroral occurrence can offer a chance for researchers to test elements of an Early Warning System. The years around 2014 arc the latest solar maximum recurring on an eleven-year solar cycle. Now is the time when aurora is more likely over populated areas, and this is the first solar inaximum with social media, an unprecedented opportunity to engage the public, the scientific community, and the media. Space weather scientists have only coarse means to predict where the aurora will occur. Forecasts derived from state-of-the-solar wind models and satellite-based observations of the Sun estimate the arrival of the coronal mass ejection material at Earth to a typical accuracy of ±8 hours. There is no way to accurately predict the strength of an incoming storm. Space weather observations are simply not plentiful enough to accurately forecast the 93 million miles of interplanetary plasma between the Sun and the Earth. However, a network of observers on the ground can provide ground truth, which could enable nowcasts. We anticipate an order of magnitude increase in accurate knowledge of where and when the aurora are visible by augmenting existing observations and models with real-time ground truth (thus going from 100s to 10s of km, and from hour to minute time-scales). Using these new data we anticipate three major impacts on geospace science: 1) Significantly more accurate, timely, and relevant methods of auroral predictions. 2) New cost-effective sources of human-derived data for observing the aurora, including volunteered geographic information, images and videos, which may aid scientists in understanding their space-based data. 3) The ability to study auroras at mid-latitudes during storms. A storm has never been imaged at mid-latitudes and the last global imaging satellites with significantly lower resolution ended their missions in 2006. Space scientists are still tillng to connect many magnetosphcric regions to their visible manifestations in the ionosphere in this data-starved, vast, and relatively young field. Other citizen science activities may yield discoveries on the shape of the dynamic auroral oval versus time and activity due to the current lacic of data on these phenomena. On October 24th, 2011 aurora was visible as far south as Alabama. Due to the excelletit timing and widespread visibility in clear skies over heavily populated areas, this storm was covered by more than 500 traditional and online news outlets from large international newspapers to blogs. In real-time, tweets exceeded one per second during the event. Additionally, thousands of Twitter messages (tweets) containing text and images documented the visible aurora in real-time (MacDonald, 2012). During the testing of our prototype, we found that the Kp index, a key measure of auroral activity, correlates strongly with the number of aurora-related tweets. This suggests that volunteered Twitter data is a valid indicator of real-world events.

UR - http://www.scopus.com/inward/record.url?scp=84974851145&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84974851145&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84974851145

VL - WS-14-20

SP - 30

EP - 32

BT - Citizen + X

PB - AI Access Foundation

ER -

Tapia AH, Lalone N, MacDonald E, Case N, Hall M, Heavner M. Aurorasaurus: Citizen science, early warning systems and space weather. In Citizen + X: Volunteer-Based Crowdsourcing in Science, Public Health and Government - Papers from the 2014 HCOMP Workshop, Technical Report. Vol. WS-14-20. AI Access Foundation. 2014. p. 30-32