NOTE: This formed the basis of our final proposal, which can be found here
SCHOME Core Team
|Baso Schomer||Decimus Schomer||Explo Schomer||Kali Schomer|
|KitKatKid Schomer||Marko Schomer||Mars Schomer||Topper Schomer|
|Faji Bing||Gaea SParker||Technomancer SParker|
1. Executive Summary
We propose an instrument, s.c.h.o.m.e. (Spectroscopy, Climate and Habitability from Orbital Measurement of Earthshine), to observe Earthshine and identify key life markers in the reflected light of the Earth. The s.c.h.o.m.e. instrument consists of an adapted visible-near infrared spectrometer and a fibreoptic cable, with the ability to monitor the spectra of light emitted from Earth. By using Earthshine to analyse the composition of the Earth’s atmosphere, we will be able to monitor O2, O3 and H2O. Results from this study will characterise habitability and life signatures of the Earth and serve as a reference when classifying the sustainability and potential life of an analogously observed exoplanet. Our study would tie in with current and future missions of international space agencies to observe exoplanets. In the long term, we could observe the effects of climate change and human influence on the Earth albedo in terms of global cloud, ice and water cover.
Our team comprises eight core team members aged 14 to 17 years old from all over the UK who are taking part in the Schome Project (www.schome.ac.uk) based at the Open University, Milton Keynes. The Schome Park Project is an education innovation research project using two private islands on the Teen Grid of the virtual environment Second Life, in conjunction with an online forum and wiki (Note the demarcation in the nomenclature; Schome refers to the educational project and s.c.h.o.m.e. refers to the proposed instrument). The Schome Park Project started on the 12th of March 2007 and although most of us were in NAGTY (http://www2.warwick.ac.uk/nagty/), we hadn’t really ‘met’ until the start of the project. The whole proposal will have been discussed and written through these mediums. The team have only ever met in real life once during the trip to meet the engineer at Surrey Satellites Technology Limited and therefore the communication has only really been through the internet through the forum and wiki, as the closed Island in teen second life has been temporarily closed. Communication through these mediums has been a critical aspect of the project. Unlike other entrants, our team have only had one opportunity to physically meet to discuss our entry. Instead, we have had to find ways to discuss and develop our ideas within Second Life and other internet resources, thus making our entry unique.
3. Scientific Background
3.1 Visible Light and the Electromagnetic Spectrum
What we see as visible light is one small part of the entire electromagnetic spectrum of radiation (see Figure 1). The frequency, the number of peaks or troughs that pass by in a given time (usually measured per second, as Hertz, Hz), describes how much energy a beam of light carries and is related to wavelength. Visible light is between 400 and 700 nm, with near infrared between 700 and 1400 nm. Our experiment would aim to measure the spectrum of wavelengths from around 400 to 800 nm, thus including visible light, in addition to some of the near infrared.
Figure 1. The electromagnetic spectrum, highlighting the wavelengths of visible light. From http://www.yorku.ca/eye/
Earthshine is the sunlight reflected back from the Earth, generally observed through its reflection onto the dark side of the moon, as illustrated in Figure 2, and so by measuring Earthshine reflected from the moon, as was our initial plan, we could observe the reflected light (albedo) spectrum of the Earth. Further study discovered that observation of reflected Earthshine was unfeasible (see section 7). Instead, we have elected to observe the Earth directly from orbit and use the spectra discovered to identify the composition of the surface and atmosphere, not only as an average but also at different locations and times. We can also monitor changes in atmospheric features and terrain such as cloud, ice and water cover of the planet. At a basic level, some terrains (such as snow) have a higher albedo than others (such as the sea). Generally, land-based topographies will reflect more light than liquid water. In addition, vegetation has a distinct reflectance pattern, with a sudden change in the reflection of light beyond a wavelength of around 700 nm. Traditionally, Earthshine is measured using ground-based telescopes observing the moon; we propose to take measurements from Earth orbit, allowing us to directly observe the Earth and decreasing interference from atmospheric effects.
Earthshine will display a number of features relating to both the atmosphere and the surface of Earth. Another major advantage of observing the Earth directly is avoiding the need to remove the the interference from the Moon (which we would have found by investigating the spectrum of light reflected from the sunlit side of the Moon). Therefore, the Earth albedo EA(λ) is more simply the light from Earth adapted to account for positions of the Earth and Sun, relative to the satellite (Equation 1). ES(λ) is defined as Earthshine and g1 is the geometrical factor which differs depending on the relative positions.
EA(λ) = (ES(λ)xg1) Equation 1
3.2.1 Previous Earthshine Measurements
There have been several projects to observe Earthshine, notably at Big Bear Solar Observatory, California, and the GOME project on the satellite ERS-2. Although some Earthshine models have been created, not enough observations have been made to create a definitive model of Earthshine which would be useful in indicating the existence of life on other planets or identifying a habitable environment. Our experiment also has the main advantage of observing the spectrum from Earth directly.
3.2.3 Features of Life Observable in light from Earth
The presence of life on Earth is indicated by features in the light absorbed or reflected by Earth's surface and atmosphere, which are observable in Earth’s albedo spectrum. These features are based on the absorption spectra of different elements, compounds and surface terrains, all of which have unique absorption bands. Identifying these spectral features in Earth albedo spectra allows us to understand externally perceivable features of life on Earth better, and this is transferable to observations of exoplanets (see section 3.3). Examples of Earth albedo spectra (observed and modelled) are given in Figure 3. Primary spectral features include:
- • O2 (630 nm, 690 nm and 760 nm) - a potential indicator of the presence of life. Without photosynthesising plants and algae, oxygen would typically be produced in low quantities. Oxygen in the atmosphere would be removed through reactions with minerals to form oxides. Without replacement via photosynthesis, the overall concentration of oxygen in the atmosphere would decrease over time. The existence of oxygen in the atmosphere of an exoplanet may therefore be an indicator of photosynthesis, and so one indicator of habitability and the presence of life.
- • H2O (700 – 720 nm) – a vital compound for life on Earth, it is likely to be important in the development of life on other planets. It is essential for many of the biochemical reactions necessary for life. It remains a liquid at a large range of temperatures, moderating climate to allow the development of life at a temperature enabling biological reactions to occur quickly. Because of its importance, planets lacking liquid water are unlikely to sustain life.
- • Vegetation Red Edge (680-730 nm) – most organisms rely directly (or indirectly) on an external energy source, and this is most often light from the local star. Any organism which takes in light will absorb high quantities of light at certain frequencies and, as an evolutionary adaptation to prevent overheating, is likely to reflect light outside these frequencies, especially in the infrared. On Earth, this is observable in terms of a rapid increase from 5% to 50% in the quantity of light reflected from vegetation beyond the “red edge”, a wavelength of around 700nm. This can be extracted from analysis of Earthshine, allowing the red edge to be studied in greater depth. However the effect of this would be small and it is therefore not a priority of the sensor to find it, as discerning such a red edge would be very difficult. If we could see this in the results, it would be a massive bonus, and would demonstrate the possibility of finding such an effect on a life supporting exoplanet.
- • O3 (broad feature over 450-750 nm) could also be monitored in relation to annual variations and human environmental impact, as its absorption band lies within the visible range. Its relatively small effect on total absorption may however prevent it from being observed.
- • By observing the spectra from Earth over a prolonged period of time it may also be possible to detect seasonal variations in spectra, such as the yearly rise and fall of CO2 caused by the larger vegetated surface area in the Northern Hemisphere, which offer evidence for the existence of life.
An exoplanet (or extra-solar planet) is a planet beyond the Solar System. To date, 276 exoplanets have been detected (www.exoplanet.eu) and most of these are giant planets that are thought to be similar to Jupiter. Several ground-based and space-based methods are used to search for exoplanets. Current space missions used to identify exoplanets include COROT, Spitzer and the Hubble Space Telescope, as well as the secondary extended mission of Deep Impact (EPOCh) which will observe the light reflected from exoplanets. Presently, Spitzer is able to provide spectral information from which exoplanet compositions can be identified. For example, water vapour and methane have been discovered in the atmosphere of planet HD 189733b, but the exoplanet itself is unable to sustain life as it is too hot (around 1000 K). Future missions include NASA’s Kepler mission (launch 2009) and ESA’s Gaia mission (launch 2011), which will both observe planetary transits (crosses) in front of the parent star to identify new exoplanets. The James Webb Space Telescope (set for launch in 2013) will observe exoplanets using infrared imaging. It will gather the infrared images and use them to characterise the age and mass of the exoplanets in addition to measuring their spectra. ESA’s Darwin mission (estimated launch 2015) is a space telescope which aims to find and study the properties and composition of Earth-like exoplanets in the infrared. Our instrument will tie in with current and future missions to observe and search for life on exoplanets. By looking at the spectra of Earth, we can characterise what makes it suitable for sustaining life, information that can be related to present and future exoplanet observations.
4. Scientific Objectives
The primary research objectives of SCHOME are to:
- 1. determine life signatures, habitability features and albedo changes related to topography and weather of the Earth from Earthshine spectra
- 2. relate our findings to current Earth climate models using the satellite’s ability to observe the Earth’s spectrum.
- 3. relate the findings from present and future exoplanet spectral studies to those of this study.
5. Scientific Rationale
s.c.h.o.m.e. allows us to observe the Earthshine spectrum and to identify what makes Earth hospitable for life by identifying the signatures of life and assessing habitability. We will be able to characterise different terrain and vegetation as each has a different albedo, and we will be able to identify signatures of life such as oxygen. During ground-based measurements of Earthshine, the weather e.g. clouds, can make the moon difficult to observe and measurements can also be affected by local atmospheric effects. Earthshine measurements taken from space, will not be affected in the same way. They will also benefit from the Earthshine having to travel through the atmosphere one less time (compared to ground to ground based observations) hence improving the signal. We aim to collect data for as long as possible (i.e. for the entire time that the satellite will be in orbit) to obtain an accurate idea of the factors affecting key features of the Earthshine spectrum. A long-term study using this satellite could potentially observe Earthshine spectral changes resulting from the impact of climate change on our planet.
This is an ambitious project and we would seek to collaborate with those from the international astronomy community to aid the data reduction process. However, we intend the identification of key Earthshine features to be conducted by our team, with additional supervision and support provided by Ph.D. students at the Open University, while a citizen science project, as detailed in the Data Analysis section, makes extensive comparisons of small scale features, which computers would find only with difficulty. We can use the results of these Earthshine observations as a reference when assessing exoplanets in terms of habitability, climate, terrain and the presence of life.
Investigating the spectrum of light emitted from the Earth requires the collection of light on the Earth facing side of the satellite with a fibreoptic cable, and the analysis of this light in the wavelengths we will be studying with a visible near-infrared spectrometer.
6.1 Visible Spectrometer
A spectrometer will split the incoming light of Earthshine by wavelength, having previously been focussed by the optical section of the instrument. The intensity of light at each wavelength will be measured and recorded by the detector. It can then be plotted on to a graph of wavelength against light intensity, which can be analysed to find a number of features. In order to pick up the key features of oxygen, water and vegetation red edge, we will need to investigate the visible and near infrared sections of the electromagnetic spectrum, with light of wavelengths between around 500 to 1000 nm. The proposed instrumentation is a miniaturised visible and near infrared spectrometer, which is a stripped down and more rugged version of the AvaSpec-2048 Fibre Optic Spectrometer. The s.c.h.o.m.e. spectrometer is based on a UV-Vis spectrometer currently being developed for the ESA ExoMars mission. Table 1 summarises key features of the core instrument, comparing the UV-Vis instrument developed for ExoMars and the proposed s.c.h.o.m.e. instrument.
|Properties||AvaSpec-2048 Fibre Optic Spectrometer||Uv-Vis Spectrometer for ExoMars 9||s.c.h.o.m.e. Instrument|
|Optical Bench||75mm focal length||75mm focal length||75mm focal length|
|Wavelength range||200-1100 nm||200 – 650 nm||400 – 800 nm|
|Resolution||0.04 – 20 nm||1 – 1.5 nm||1 – 1.5 nm|
|Sensitivity||5000 counts/µW per 2ms integration time||5000 counts/µW per 2ms integration time||5000 counts/µW per 2ms integration time|
|Detector||CCD linear array, 2048 pixels||CCD linear array, 2048 pixels||CCD linear array, 2048 pixels|
|Power||1.92 W||200 mW||200 mW|
|Dimensions||175 x 110 x 44 mm||100 x 130 x 30 mm||Reconfiguration of the ExoMars instrument allows us to comply with competition constraints|
|Mass||719 g||~100 g||~100 g|
I believe this needs to be updated
Due to mass and space constraints, a pin-hole camera with an additional lens is proposed to focus Earthshine and Moonlight to the spectrometer. The camera consists of a tiny hole (pin-hole) that focuses incoming light from a wide field of view to the lens, which then focuses the light to the spectrometer.
Figure 4. Pin-hole camera with a lens focusing light into the spectrometer
7.1 Signal to Noise
7.1.1 Observing the Moon
In order to investigate the feasibility of our project, we needed to calculate the size of the signal, and the quantity of noise which will interfere with this signal. The rate at which the Sun's surface emits light was found to be 6.3*107 Wm-2, which means that, the Sun's surface area being around 6*1018 m2, the total power of the light released is about 3.8*1026 W. At the Earth's distance from the Sun of around 1.5*108 km, this is spread over a spherical area of 2.8*1017 km2, of which the area of the Earth facing the Sun takes up approximately 1.3*108 km2. The quantity of light from the Sun the Earth receives is therefore 2.2*109 times less than that emitted by the Sun, and so equal to 1.7*1017 W. As the average albedo of the Earth is approximately 0.367, the quantity of light reflected from Earth is 6.2*1016 W. At the moon's distance from the Earth (an average of 384 Mm), this is reflected in a hemisphere of 9.3*1011 km2, meaning that the moon, with a reflecting area of 9.1*106 km2, receives 9.8*10-6 of the light reflected from Earth, or 6.1*1011 W. The moon, with an albedo of 0.12, reflects 7.3*1010 W, which, at the distance of the satellite (approximated to the Earth/moon distance of 384 Mm), is in a hemisphere of 9.3*1011 km2, or 9.3*1017 m2. The collection area, estimated at 1 m2, would therefore receive 7.8*10-8 W. With the light we will be looking at, with a wavelength of around 700nm, each photon carries 2.8*10-19 J, so the signal would consist of 2.8*1011 photons per second. Not all of this will be in the range of wavelengths our sensor is capable of detecting, but as approximately half of the Sun's light is between 500 and 700nm in wavelength, the signal size would be large enough to be useful.
The main sources of noise which would interfere with this signal are the light from stars other than Sol and the light from other planets, which will have different features from Sol's light and so could confuse results. The total light from other stars can be calculated from the number of stars and their apparent magnitude from Earth in relation to that of the Sun. Using the frequency with which different magnitudes occur from http://www.nso.edu/PR/answerbook/magnitude.html, the total light from other stars at the Earth is the quantity of light from the Sun * 10-8. A detector facing the Sun would take in 1.4*103 W of light from the Sun alone, and thus a 360o field of view would take in 1.4*10-5 W of starlight (ignoring the Earth's shadow). In order to ensure that the signal can be detected among the noise, keeping a signal to noise ratio of at least 10:1, the noise cannot comprise of more than 7.8*10-9 W. This would limit the field of view to 5.6*10-4 times the area of the full 360o sphere. The circumference of the full sphere being 360, thus the radius 180/pi, its surface area is 129600/pi. This means the field of view could at maximum have an area of 73/pi, and so form a square with dimensions of 4.8o in each direction. As section 7.2.1 shows, this is too small to be feasible, making it necessary to observe the Earth directly.
7.1.2 Observing the Earth
Using the previous, now defunct, calculations, we know the quantity of light reflected from Earth to be 6.2*1016 W. At the satellite's distance from the centre of the Earth, of around 7 Mm, this light would be spread over 314 Mm2, or around 3.1*1014 m2. The intensity of light from the entire Earth, taking up a field of view of around 120o, would therefore be 200 Wm-2. The collector, with an assumed area of 1 m2 and field of view of 30o, would pick up 1/16 of this, so 12.5 W. With the light we will be looking at, with a wavelength of around 700nm, each photon carries 2.8*10-19 J, so the signal will consist of 2.1*1019 photons per second. Again, not all of this will be in the range of wavelengths our sensor is capable of detecting, but as around half of the Sun's output is in the observed range, the signal size will still be large enough to be useful.
Again using the previous calculations, the light from stars at Earth is 1.7*109 W, which means that around 8.5*108 W of light from other stars reach each hemisphere, so affect the signal. As this is around 2*108 times less than the intensity of the light reaching Earth, stars other than Sol will not interfere greatly with the signal. The satellite's orbit being sun synchronous with a fixed time during daylight hours, only the light reflecting from those planets closer to the Sun than Earth will affect the signal. Venus, with a maximum apparent magnitude of -4.6, and Mercury, with an apparent magnitude of -1.9, compared to the Sun's apparent magnitude of -26.7, will have a maximum brightness equal to the Sun's divided by roughly 2.522.1 and 2.524.8 respectively, or the Sun's brightness * 1.6*10-9 and * 1.4*10-10, respectively. This means that the light at the Earth from both is, at maximum, about that of the Sun at Earth * 1.7*10-9, or 2.9*108 W. As this is clearly far smaller even than the noise from other stars, planets will not have an effect.
7.2 Observation Time
7.2.1 Observing the Moon
In order to know the feasibility of observing the Moon in more detail, we must find the length of time for which observations of Earthshine are possible. First, the total length of time in which the entire moon is in shot, regardless of what proportion of this time is useful for our measurements, is dependent on both the amount of time for which the moon is in shot per swathe, and the number of swathes per day which would include the moon, which are dependent on the 'vertical' (roughly North/South) width of field of view, labelled V and the 'horizontal' width of field of view, labelled H. For ease of calculation, we will assume a rectangular view. It is known that the moon has an apparent width of 0.5 degrees from Earth, and the altitude being neglible when compared to the distance between the Earth and moon, we shall use this value.
In each swathe which contains the moon in its horizontal scope, the moon is visible from (V-0.5)/2 degrees above the line from Earth to moon to (V-0.5)/2 degrees below, hence for an arc of V-0.5 degrees. The orbital period being approximately 100 minutes, or 1.67 hours, the satellite orbits at roughly 216 degrees an hour, making the number of hours per useful swathe (V-0.5)/216. Because the moon has a maximum inclination of less than 29 degrees, that is it is never more than 29 degrees above or below the equator, almost the same method can be used for the number of useful swathes per day. First, the moon takes 27.3 days to complete a full orbit, so in a day moves 13.18 from its previous longitude. As the moon's inclination is less than 90o, it progresses relative to the Earth, which means the satellite takes roughly 1.04 days (1498 minutes) to reach the same longitudinal position relative to the moon, rather than the 1 day to reach the same longitudinal position relative to the Earth. The satellite therefore makes 1498/100=~15.0 latitudinal orbits per 'day' relative to the moon, making swathes 24.96 degrees away from each other, centre to centre. As the moon is in shot for H-0.5 degrees, it will be in roughly (H-0.5)/25.0 swathes per 1.04 days, so approximately (H-0.5)(V-0.5)/5400 hours per day. Given a maximum operational lifetime of 8 years, or 2922 days, the moon will be in shot for a maximum of roughly (H-0.5)(V-0.5)/1.85 hours.
Our previous signal to noise calculations have given a maximum possible field of view of 4.8 degrees in each direction, which leaves a maximum possible time in which the moon will be in shot of 4.32/1.85=10 hours. This figure is before accounting for the time in which observing the moon will be useful, which will likely divide the time by at least 20, and yet is still too small to justify the experiment. Clearly, we need to find a solution to increase the time available for observation. Possibilities are listed below:
- Use more than one spectrometer/collection area-cumulatively they will have a greater time period available, as the moon can be in the shot of one but not the other
- This is feasible in terms of space and budget, as the individual spectrometers/collection areas are inexpensive and take up relatively little room
- Objection: at most the observation time can be extended to 2 or 3 times the length. This increase is almost certainly insufficient.
- Allow the collection area to move-it would 'track' the moon along a single plane of rotation, across each swathe, giving more time
- As the main constraint is on the size of the field of view, this allows the time to be increased greatly
- Objection: this adds a tremendous amount of complexity and cost to the experiment, unfeasibly so
- Observe the Earth directly-as the Earth is both larger and much, much closer the field of view would only contain reflections from Earth, thus reducing the noise greatly
- This would also increase the size of the signal, by removing the extra distance to the moon and back and the moon's low albedo
- Objection: Much would have to be altered in the proposal
Clearly, the objection to the last point is far smaller than the fatal flaws in the other two proposals, and so observing Earthshine directly is the path the experiment shall follow.
7.2.2 Observing the Earth
As the satellite is using a sun synchronous orbit, it will always be observing points on Earth at the same time of day for half the time, and at the same time of night for half the time, meaning that for half the time there will be sufficient light for the spectrometer to function, as shown in the signal to noise section 7.1.2. This gives a total of at least 21912 hours of observation, in a minimum 5 year lifetime. During this time it will observe almost every point on the Earth's surface (the poles may be outside the most extreme latitudes of orbit, as the inclination is less than 90o). Using a 100 minute period, the satellite makes 14.40 orbits per day, returning to the same longitudinal position above the Earth, with swathes cutting exactly 25o centre to centre. The satellite therefore takes 5 days to return to its original position (both longitudinal and latitudinal), having made 72 orbits in that time. In each day's cycle the satellite regresses by 0.6 of an orbit (10o of longitude, 216o of latitude). In the full orbital cycle of 5 days, an orbit is made along every 5 degrees of longitude. At least 365 such full cycles will be completed in the satellite's operational lifetime.
The satellite orbits at an altitude of roughly 700 km. Using figure 5, this makes the total distance satellite to Earth's centre roughly 7070 km. Using the sine rule, we know therefore that sin(A)/7070=sin(θ)/6370, where θ is half the width of the field of view and A is the angle between the edge of the field of view and the intersecting radius, hence
A=sin-1(7070sin(θ)/6370). As A must be greater than 90o (B being 180o-90o-θ>=75o (θ being <=15), and as length c>b, C>B, giving A>150o), angle Θ, half the width of the swathe, is 180-θ-(180-sin-1(7070sin(θ/2)/6370))=sin-1(7070sin(θ)/6370)-θ. At the maximum value of θ as half 30o, Θ is therefore 1.694o, making the swathes each 3.388o wide.
With the 5o spacing of the orbits within a full orbital cycle, not every point on the Earth's surface is observed, and instead just above two thirds of the surface within ±83.694o of latitude is observed (with the poles outside than the 98o inclination's scope). As the satellite orbits at 3.6o a minute, each point on the swathe is in shot for 0.94 of a minute per orbital cycle, so is observed for at least 343 minutes (~5.7 hours) in the satellites operational lifetime, which gives more than enough information on the areas observed.
7.3 Data Acquisition
The spectrometer we will be using takes 1024 readings from each portion of the spectrum at 16 bit every half a second, meaning that a full measurement is composed of 2048 bytes, or 2 kB per 0.5 seconds. As each measurement takes 0.5 seconds, and the satellite orbits at 216o an hour, or 0.06o a second, each measurement will cover an area only very slightly larger than the field of view, ie 3.388o of the Earth's circumference in each direction, or roughly 1902=36100 km2, which is small enough to distinguish between different terrain types, yet should still average the spectrum sufficiently.
With the limit of 1 MB per day, a maximum of 512 measurements, or 256 seconds of measurement, are possible per day. As only half of the satellite's orbit is above the dayside, and thus useful, these 512 measurements are split between 43200 seconds, thus separated (centre to centre) by 84.375 seconds each, during which time the satellite will orbit just over 5o, ensuring substantial latitudinal overlap between measurements. A measurement is taken at the same time of day each day, and thus at the same place each cycle, allowing easy comparison between the at least 365 measurements that shall be taken each of the same portion of the Earth' surface.
7.3 Fields of View
In order to allow the experiment to include data from as large an area of the Earth's surface as possible, allowing us both to find average spectra representative of the surface of the Earth and to look at any particular region of the Earth's surface, we shall use the largest field of view possible with the optics we have chosen. A simple fibreoptic cable will allow us to observe a field of view of around 30o.
This size of field of view will allow, as calculated in previous sections, observation of the majority of the Earth's surface. At the same time, the surface area covered in a single field of view will be small enough to allow differentiation of different surface types. This allows us to, through analysis of the data and mapping of the area of the Earth covered by each measurement, observe the effect of each type of surface area on results. This increases the likelihood of observing such effects as the red edge, which will, if observable at all, be so on observation of the Amazonian region.
8.1 Mass Budget
8.2 Power Budget
8.3 Financial Budget
The financial budget will be kept within competition constraints for finance as the off the shelf price for the spectrometer is £2000, this is the most expensive component of our instrument. Any fibre optic cables etc will be well covered by the remaining £98 000.
9.1 Data Analysis
9.2 Use of results
9.2.1 Citizen Science
Citizen science is the term that is used to describe projects of scientific work where members of the public help to analyse the data obtained through a scientific research project. The use of citizen science allows scientists to accomplish feasible research goals that would be otherwise difficult to achieve. It allows many people to view and analyse the data obtained, meaning that the same piece of data will be examined by numerous individuals. In addition to this, citizen science actively engages the public and involves them in up to date scientific research. The general public has little or no knowledge of what current scientific research is being carried out today, so feel distanced from science and have no interest in it. This is also shown by the decline of students opting to study a science after GCSE or in higher education. Citizen science aims to make the general public get involved in research and to give them a general and informal education in aspects of science.
There are a few projects currently being held that involve citizen science, such as World Water Monitoring Day, NASA’s Starduct@home and the Clickworks, as well as various projects run by the Cornwall Laboratory of Ornithology and the long running Christmas Bird Count by the Audubon Society (Started in 1990). Inspired by the Stardust@home project, GalaxyZoo is an online project where members of the public assist in the classification of over a million galaxies. The volunteers have to identify if a galaxy is elliptical or spiral and, if spiral, whether they rotate clockwise or anti-clockwise.
As of August 2007, more than ten million galaxies had been classified by over one hundred thousand volunteers (http://arxiv.org/PS_cache/arxiv/pdf/0804/0804.4483v1.pdf). By combining the classification of more than one hundred thousand participants, GalaxyZoo have been able to produce catalogues of galaxies that agree with professional astronomers to an accuracy of greater than ten percent. One of the largest catalogues in GalaxyZoo has over three hundred thousand galaxies that have been reliably classified by the general public. As each galaxy is classified by multiple volunteers, there is a lower chance of errors occurring.
There are many advantages to this means of analysing data. According to Kevin Schawinski (a member of the team behind the project), a human brain is much better and more reliable at pattern recognition tasks than a computer program is, so the human element in the classification of galaxies is vital. Without the vast number of volunteers participating in GalaxyZoo, it could take years for all of the images of galaxies to be classified by professional astronomers. This is the method we propose to analyse the bulk of the data obtained. In addition to the vast numbers of classifications which this approach can make, the human participants can use their intuition and extensive online training to make classifications which computers would find difficult. In this way an online system, particularly using the resources available to us in Second Life, would use thousands of people to compare the Earthshine spectra collected with expected signatures, and compare the spectra collected at different times and places, picking up on the differences which terrain make, as well as seasonal variations, and hopefully analyse changes across the satellite's lifetime. After completing online training on the signatures they could expect to find, and using already known spectra to confirm their accuracy, a checklist would allow them to select which substances appear to be in the spectrum and highlight unusual formations to pass onto a PhD team. Likewise, checklists would allow them to compare the relative predominance of certain formations or substances. In order to aid this, Google Earth (http://earth.google.com/)can be used to display the section of the Earth the spectrum is taken from.
We envisage a website that any member of the public can access. We would use this to allow the general public to participate in the analysis of the data obtained from our experiment. Using GalaxyZoo as a guide, we would create the website so that the public would sign up and then have to go through tutorials to ensure that they know what they are doing and for us to be sure that they can correctly interpret the data and classify it.
We visualise our website to be simple enough for the public to use without having any specialist scientific knowledge. As well as the classification of the spectra, we could have other activities etc to motivate the public to participate, such as leader boards for the most spectra analysed and games for younger participant to gain the widest range of volunteers possible.
To generate interest, and ultimately participation, for the public in the website, we would use the media attention that the competition itself generates. The fact that all the members of the team are from different parts of the country and are all at different stages in education would also be advantageous for promoting participation in the project. We could use the connections and education networks in our various areas and develop a catalogue of students from all over England that would participate in this project through their own school. This has the added bonus of encouraging students to actively participate in scientific research and demonstrates to them what science can do for the general public and how anyone can be involved. This also promotes the study of science subjects beyond GCSE as students get a glimpse of what scientific research is like for actual scientists.
The web-based citizen science project we envisage creating to analyse the data from our instrument, will be interfaced to Second Life. Second Life (r) is an online virtual community in which users (also known as avatars / residents) create content for a number of reasons, eg for educational or recreational purposes or for e-commerce (where residents can buy and sell objects or virtual land). There are two parts to Second Life; the Main Grid for adults and the Teen Grid for those aged 13-17 years old. At the time of writing there were a total of 14,856,193 combined residents on the Main and Teen Second Life Grids (though only around 850,000 of these are active in each month(http://secondlife.com/whatis/economy_stats.php)), with up to around 60,000 online at any one time (http://taterunino.net/peakmonthlyconcurrency. jpg).
The Open University holds a strong presence in Second Life (both on the Main and Teen Grids), all students and staff involved in this proposal are participants in the Schome Park project (resident on the Teen Grid) whilst there is a complimentary project for adults(Schomebase) present on the Main grid. Both Schomebase and Schome Park could serve as a base to implement a citizen science project through Second Life in addition to the traditional website based scheme. Consequently Second Life users can participate in data analysis from this mission. Working groups or focussed discussions regarding the data can be set up, thus feeding into the educational framework currently being developed in Second Life by various Open University research groups. In addition to actual data analysis, spin-off activities such as planetary/Earth science lectures can be given in-world, replicas of the S.C.H.O.M.E. instrument can be built and scale models of the Earth including changing albedo over the life time of the mission can be created.
There currently exist a few planetary and space science educational projects within Second Life. These include: - a detailed scale model of the International Space Station on the Eye4You Alliance island - NASA have an amphitheatre in the Main Grid where NASA tv is streamed into Second Life - there is an International Space Museum on the Main Grid (http://www.spacetoday.org/SpcStns/SecondLife/SL_ISS.html). To the best of our knowledge no other group is using the 'general public', Second Life user, in the manner we propose - relying instead on specialists within the individual organisations to deliver science talks. With a community of almost 15 million people, a citizen science project in Second Life could truly engage and inspire the public with respect to planetary/Earth science by allowing them to be involved in data analysis from a real life mission!
Our team, consisting of eight core team members from different parts of the UK, has developed a proposal for this competition within the virtual world of Second Life, using the Schome Park Project forum and wiki to organise our ideas. We propose an instrument, SCHOME (Spectroscopy, Climate and Habitability from Orbital Measurement of Earthshine), to measure the reflected light of the Earth (Earthshine). The SCHOME instrument consists of an adapted visible-near infrared spectrometer and a pin-hole camera and has the ability to take 186880 readings per year over the lifetime of the satellite, each reading from an area roughly the size of Belgium, with readings taken every 5o of the satellite's orbit, with a total of roughly two thirds of the Earth's surface observed. The main objectives of the instrument are to 1) determine life signatures, habitability features and albedo changes related to topography and weather on the Earth and 2) relate these results to current Earth climate models as well as present and future exoplanet spectral data.
Kaiser, P. York University. http://www.yorku.ca/eye/
Meadows, V., Crisp, D. and Tinetti, G. Characterization of Terrestrial Planets From Disk-Averaged Spectra: Spatially and Spectrally Resolved Planetary Models. http://nai.arc.nasa.gov/team/customtags/projectreports.cfm?teamID=57&year=7&projectID=1385