Exploring Descent Camera Designs for Planetary

Penetrators

Name: George Cann.

Department of Space & Climate Physics, University College London, United Kingdom.

Abstract

A planetary penetrator is a robust, instrument dense package that impacts the surface of

a planet or moon at speeds of ≈ 300 ms−1. This reports reviews the scientific and instrument

design requirements of a theoretical descent camera (DC) for a Europa penetrator mission.

The report begins with a literature review of Europa, ESA’s CLEO/P Concept, NASA’s Lunar

Reconnaissance Orbiter WAC, NASA’s Europa Mission EIS and ESA’s JUICE mission JANUS

instrument.

The report then presents a set of theoretical models relating to the DC’s operation and

design, including its operational wavebands, imaging resolution, mass constraints, data volume

constraints and power constraints. Theoretical models of the DC’s operation are then written

and simulated using Matlab.

The report then presents a set of results, which includes times during the DC’s operation

when its spatial resolution exceeds the best resolution of JANUS (≈ 10 m/pixel) and the

EIS’s NAC (≈ 0.5 m/pixel). Finally the report analyses the results and concludes that the

addition of a DC, based on the theoretical model, is potentially of scientific benefit for a

Europa penetrator mission.

1

Contents

1 Introduction 3

2 Literature Review 3

2.1 Europa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.1.1 Europa’s Orbit and Rotation . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.1.2 Europa’s Surface Composition . . . . . . . . . . . . . . . . . . . . . . . . . 4

2.1.3 Europa’s Surface Structure & Geomorphology . . . . . . . . . . . . . . . . . 4

2.1.4 Europa’s Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.1.5 Jupiter’s Magnetosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.2 ESA’s CLEO/P Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.3 NASA’s Lunar Reconnaissance Orbiter - WAC . . . . . . . . . . . . . . . . . . . . 7

2.4 NASA’s Europa Mission - EIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.5 ESA’s JUICE Mission - JANUS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

3 Methods 11

3.1 Descent Camera Design Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

3.2 Descent Camera Design Theory - Matlab Simulation Method . . . . . . . . . . . . 13

3.3 Wavebands Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.4 Imaging Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.5 Mass Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.6 Power Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.7 Data Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

3.8 Digital Terrain Model Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

4 Results 17

4.1 Descent Camera Design Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17

4.2 Wavebands Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

4.3 Imaging Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4.4 Mass Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

4.5 Power Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.6 Data Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

4.7 Digital Terrain Model Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

5 Discussion 27

5.1 Descent Camera Design Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

5.2 Wavebands and Imaging Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

5.3 Mass, Power and Data Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

5.4 Digital Terrain Model Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

6 Conclusions 29

6.1 Future Research and Improvements . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

7 Acknowledgments 30

8 Bibliography 30

2

1 Introduction

This report aims to review the scientific and instrument design requirements of a DC, which is

to be carried on-board a planetary penetrator or a payload delivery system (PDS), for a Europa

penetrator mission. A planetary penetrator is a robust, instrument dense package, that impacts

the surface of a planet or moon at speeds of several hundred metres per second. Once buried in

the planet or moon, a penetrator is capable of performing in-situ measurements of the planet or

moon’s sub-surface, such as chemical composition measurements. A DC is capable of providing

important information regarding the exact location of a penetrator’s impact site and the nature of

its surrounding terrain. This report aims to answer the following questions:

1. What are the scientific goals of the DC?

2. What is the most suitable type of imaging detector for the DC?

3. What are the mass, power and data constraints on the DC?

4. What is the maximum science return given the above constraints?

5. What is the ideal DC imaging resolution, wavebands, imaging heights and imag-

ing frequencies?

6. What are the ideal image exposure lengths, bits per pixel and data rates?

7. What does the DC’s theoretical field of view (FOV) look like?

The method of inquiry in this research project has been through a literature review and the

development of a set of theoretical designs for an unconventional DC. The need to downlink the

DC’s images before the DC’s destruction on impact, the need to obtain multi-colour panoramic

images and the need to function on a rapidly spinning platform all contribute to its unconventional

design.

2 Literature Review

Recent evidence suggests that Europa hosts a H2O(l) based ocean between its outer icy crust and

the outer surface of its outer mantle. This has sparked scientific interest in Europa’s potential

habitability. Interest in the Jovian system is evidently high, with the recent arrival of NASA’s

Juno spacecraft at Jupiter on 4th July 2016, NASA aiming to launch its Europa Mission in 2022

and ESA aiming to launch its JUpiter ICy moon Explorer (JUICE) mission in 2022.

This section presents a literature review of current research on Europa, Europa spacecraft

cameras and Europa mission concepts, with the sources of the literature review being predominantly

research papers and textbook material. The literature review has been used as an essential guide

in the design of the DC.

2.1 Europa

This subsection provides a background overview of Europa and provides context for the devel-

opment of the DC’s design. The subsection covers Europa’s key parameters including its orbit,

rotation, surface composition, surface structure, geomorphology, atmosphere and the influence of

Jupiter’s magnetosphere on the Europan environment.

2.1.1 Europa’s Orbit and Rotation

Europa is one of four Galilean moons that orbit Jupiter and has a semi-major axis a = 6.7108 ×105 km and orbits Jupiter in 3.5518 days. Europa has a mass of 4.797±1.5 ×1022 kg and a radius

of 1565 ± 8 km [3]. Key properties of all the Galilean moons are displayed below in Table 1.

3

Figure 1: Final view of the Jovian system taken by NASA’s Juno spacecraft on 29th June 2016, showingall four Galilean moons. [22]

Table 1 - Galilean Moon Properties [3]

Satellite a (103

km)

Period

(days)

i (deg) Mass

(1020 kg)

Radius (km) Albedo

Io 421.77 1.769138 0.04 893.3± 1.5 1821.3± 0.2 0.61

Europa 671.08 3.55181 0.470 479.7± 1.5 1565± 8 0.64

Ganyemede 1070.4 7.154553 0.195 1482± 1 2634± 10 0.42

Callisto 1882.8 16.689018 0.28 1076± 1 2403± 5 0.20

Orbital resonance occurs between the Galilean moons. Io orbits Jupiter at twice the rate

of Europa, similarly Europa orbits Jupiter twice as fast as Ganymede. Furthermore conjection

between Io and Europa always occurs at Io’s perijove. [3] All of the Galilean moons have a

synchronous rotation, however Europa’s outer icy crust may rotate faster than synchronous, on

the order of 1 extra rotation every 50000 years [3]. This report will not address the orbital

characteristics and trajectories of Europa and a penetrator mission, however it is likely that they

will influence the choice of the penetrator’s landing site.

2.1.2 Europa’s Surface Composition

In Calvin et al. (1995) [9], the spectra of Europa’s surface, from Earth based observatories, is

compared against a laboratory scattering model of 100 µm H2O(s) ice crystals, Figure 2 (right).

The close agreement of the laboratory spectra’s shape in the 1 - 2.5 µm range and Europa’s surface

implies that H2O(s) is Europa’s main surface constituent. [9]

Furthermore comparing the spectra in Calvin et al. (1995) with the combined Visible/NIR

spectra of Europa’s surface, observed by NASA’s Galileo Solid State Imager (SSI) and the Near

Infrared Mapping Spectrometer (NIMS) [11], Figure 2 (left), shows that both are strongly in agree-

ment. This further validates the finding that Europa’s surface composition is primarily H2O(s).

Furthermore both spectra show that Europa’s surface has a relatively high reflectance at visible,

near-infrared (NIR) and mid-infrared (MIR) wavelengths. NIR corresponds to wavelengths of ≈0.78 µm to 3.0 µm and MIR corresponds to wavelengths of ≈ 3.0 µm to 50.0 µm. This suggests

that these are the ideal wavebands for the DC to observe.

2.1.3 Europa’s Surface Structure & Geomorphology

Characteristic impact craters suggest that Europa’s outer icy shell is approximately 20 km thick.

This thickness is consistent with convective upwelling estimates of warm ice in diapers. The process

of convective upwelling is thought to form Europa’s characteristic dark spots called lenticulae. [3]

Chaos terrain is thought to be caused by H2O(l) from Europa’s sub-surface ocean rising up

through its outer icy crust, an example of chaos terrain is shown in Figure 3. The thickness of the

4

Figure 2: Galileo SSI and NIMS combined visible and NIR spectra of Europa, J.C. Granahan et al. (1997)(left). Laboratory 100 µm H2O(s) spectra against Europa spectra (right). [9] [10]

Figure 3: Conamara Chaos region on Europa. [4]

shell at Conamara Chaos is though to be less than 6 km thick. [3] Europa’s surface is known to

be heterogeneous and chaotic on a decameter scale and is likely to be so at smaller spatial scales,

Greeley et al. (2009). The chaotic terrain makes landing a spacecraft on the surface of Europa

challenging and risky. Greeley et al. (2009) suggests that images at the scale of 1 m/pixel will be

critical for characterisation of future Europa landing sites. Although Europa’s surface is chaotic

Europa has a relatively low global variation in elevation (≈ 400 m). This may be advantageous

in creating Digital Terrain Models (DTM) around the penetrator’s impact site from stereoscopic

imaging.

Europa has a mean density of ρ = 3.01 g cm−3, which is indicative of a rocky icy composition,

furthermore Europa’s moment of inertia ratio IMR2 = 0.346 implies that Europa is differentiated

and centrally condensed. [3] The interior structure of Europa is most likely an outer icy crust, a

sub-surface H2O(l) based ocean, a rocky outer mantle and an inner iron metallic core. [3]

Europa has very few impact craters, which is likely due to crater removal and impact frequency.

This also suggests that Europa’s surface is relative young, on the order of tens to at most a few

hundred million years.[3]

5

2.1.4 Europa’s Atmosphere

Europa’s surface is covered by a H2O(s) based crust and a tenuous oxygen atmosphere, with an

atmospheric surface pressure psurface = 0.1 µPa. [3] A process of sputtering emits H2O(g) molecules

from the surface that subsequently photo-dissociate into hydrogen and oxygen. Sputtering is

defined as the continuous bombardment by high energy particles from sources such as the solar

wind and magnetospheric plasmas, that causes atoms and molecules from a celestial body’s surface

to be ejected.[3] Due to Europa’s relatively low surface gravity gEuropa = 1.315 ms−2 hydrogen

escapes, leaving the atmosphere dominated by oxygen. The lack of a substantial atmosphere

around Europa is advantageous in the design of a penetrator mission, as the effects of atmospheric

drag during descent on a penetrator and a PDS will be negligible.

2.1.5 Jupiter’s Magnetosphere

Europa orbits within Jupiter’s magnetosphere in a high radiation environment, where charged

particle intensities near Europa’s surface range from 1 keV to tens of MeV, Paranicas et al. (2008).

It is possible that spacecraft anomalies could be induced in a penetrator and a PDS from the intense

Jovian radiation environment. Given the relatively high radiation environment it is possible that

radiation hardened instruments may be preferential [17], such as the JANUS instrument selected

for ESA’s JUICE mission. [18]

On the 3rd January 2000 NASA’s Galileo spacecraft performed a flyby of Europa. From this

flyby Galileo’s on-board magnetometer measured fluctuations in Jupiter’s magnetosphere. These

fluctuations could be explained if a planet-scale liquid ocean lay beneath Europa’s outer icy crust,

Kivelson et al. (2000). This magnetometer data is the most convincing evidence of the existence

of a Europan sub-surface H2O(l) based ocean. [3]

2.2 ESA’s CLEO/P Concept

Figure 4: CLEP Penetrator and PDS design. [1]

NASA’s Europa Mission, formerly known as Clipper, is currently a Phase A mission to Europa,

which NASA has invited ESA to contribute to. If found scientifically valuable, meaningful and

financially viable, ESA’s contribution could become part of ESA’s Cosmic Vision programme. [1]

ESA’s Future Mission Office envisioned three mission concepts for an ESA contribution to

NASA’s Europa Mission, these include an Io minisat concept Clipper ESA Orbiter Io (CLEO/I), a

Europa minisat concept Clipper ESA Orbiter Europa (CLEO/E) and a Europa penetrator concept

Clipper ESA Penetrator (CLEP). [1]

CLEP consists of two stages, a penetrator and a PDS, the design of the CLEP penetrator and

PDS are shown in Figure 4. [1]. The penetrator and PDS are designed to impact the surface of

6

Europa at ≈ 300 ms−1. [1] [7] Only the penetrator is designed to survive the impact. After impact

instruments on-board the penetrator are designed to take measurements of Europa’s sub-surface.

CLEP’s primary science objective is to perform astrobiological science of the surface and sub-

surface environment, which could potentially be performed through gas chromatography and mass

spectrometry, Greeley et al. (2009). The secondary science objective is geophysical and includes

confirming the existence and depth of Europa’s sub-surface H2O(l) based ocean, as well as charac-

terising the variation of Europa’s properties with depth. [6]

2.3 NASA’s Lunar Reconnaissance Orbiter - WAC

NASA’s Lunar Reconnaissance Orbiter (LRO) was launched on 18th June 2009 with the primary

goal of determining key properties of the Moon that will enable future human lunar exploration.

LRO’s Lunar Reconnaissance Orbiter Camera (LROC) is a camera system built by Malin Space

Science Systems (MSSS) and consists of three main components, which include:

1. Wide Angle Camera (WAC).

2. Narrow Angle Camera (NAC).

3. Sequence and Compressor System (SCS).

The main element of LROC, that is of relevance to the design of the DC is its Wide Angle

Camera (WAC). LROC’s WAC implements a pushframe imaging method and has significant her-

itage from the MARCI instrument on-board the Mars Reconnaissance Orbiter (MRO), Bell et al.

(2009). LROC’s WAC consists of the following two main components:

1. WAC Optics. This includes a visible lens, a ultraviolet (UV) lens, a prism and a Color

Filter Array (CFA), the set of filters that LROC’s CFA consists of are shown in Table 2.

2. WAC Electronics.

Table 2 - LROC Filters

Filter Central

Wavelength

(nm)

Purpose

#1 (UV) 321 Ilmenite minerals.

#2 (UV) 360

#3 (Visible) 415 Variation in Fo-number (MgFe ratio) of

olivine.

#4 (Visible) 566

#5 (Visible) 605

#6 (Visible) 643

#7 (Visible) 689

As LROC moves in its lunar orbit its framelets sweep across the Moon’s surface, building up

narrow long image swaths. A framelet is defined as a 14× 1024 pixel strip of a detector. LROC’s

pushframe imaging method involves imaging, at a rate that allows images from its framelets to

overlap. Operating in colour mode allows LROC to obtain continuous coverage in 7 different

colours without the need for a filter wheel mechanism. The ability of LROC to obtain multi-colour

images whilst avoiding the use of a filter wheel mechanism is a favourable attribute, as spacecraft

mechanisms are common points of failure.

LROC’s WAC, shown in Figure 6 (left), is capable of monochrome and colour imaging, in

monochrome mode only the 605 nm framelet is used, whilst in colour mode all 7 framelets are

used. It may be beneficial for the DC to be capable of operating in both monochrome and colour

modes. The choice of being able to operate in both modes and when to operate in each mode,

could play an important role in data volume reduction. LROC’s WAC has a 92◦ monochrome

7

Figure 5: LROC pushframe monochrome (left) and colour images (right), Robinson et al. (2010). [15]

FOV, a 61◦ visible FOV and a 59◦ UV FOV. Furthermore LROC’s lenses are fused such that they

are optimised for wavelength range of its CFA. This suggests that the DC visible, NIR and MIR

lenses should be fused such that they are optimised for the wavelength range of its filters.

LROC uses a SCS, shown in Figure 6 (right), to transfer commands from the LRO bus to the

WAC and NAC, so as to acquire image data, this suggests that an SCS equivalent will be required

for the DC. Moreover based on an LROC type design the optical elements for the DC should

include a visible lens, a NIR lens, a MIR lens, a prism and a CFA.

Figure 6: LROC WAC (left) and SCS (right), Robinson et al. (2010). [15]

From Robinson et al. (2010) it is apparent that the filters used in the CFA correspond to

spectral signatures of minerals that are related to the scientific objectives of the LRO. This suggests

that the filters used in the DC, should be suited to the detection of minerals and molecules, that

correspond to the scientific objectives of the DC.

LROC observes a relatively low intensity UV signal. From the spectra in Figure 2 (left) it can

be seen that the NIR and MIR signals observed by Galileo are also relatively low. It can be seen

that peaks in the NIR occur at ≈ 1.7 µm and 2.2 µm and that their scaled reflectance is ≈ 0.45.

In LROC, UV signals are recorded by summing a set of pixels. Therefore if the DC camera is to

record NIR and MIR signals at Europa, it may also need to sum a set of pixels. The DC may also

8

require visible, NIR and MIR baffles, so as to block light coming from sources outside the FOV of

the DC, such as Jupiter, Jupiter’s other moons and the Sun.

2.4 NASA’s Europa Mission - EIS

NASA’s Europa Mission is a multiple flyby mission to Europa, based on the Europa Clipper mission

concept and is expected to be launched in 2022. Questions that the EIS aims to answer include:

Where are the scientifically compelling landing sites for a potential lander? And what is the nature

of the surface and the potential landing hazards, at the scale of a potential lander? The Europa

Imaging System (EIS) is one of nine instruments proposed for NASA’s Europa Mission. The

EIS combines a Narrow Angle Camera (NAC) and a Wide Angle Camera (WAC) to investigate

Europa’s geology, composition, outer icy shell and sub-surface ocean.

Figure 7: Preliminary EIS NAC and WAC design. [23]

Elements of the EIS that are of particular interest in the the design of the DC is the NAC, the

6 broadband filters and the CMOS detectors.

The NAC has a 2.3◦ × 1.2◦ field of view (FOV) and a 10 µrad IFOV. [17] From 50 km above

Europa’s surface the NAC is has a resolution of 0.5 m/pixel over a 2 km wide swath. [17] The

ability to of the NAC to gimbal allows for > 95% global coverage of Europa’s surface at ≤ 50

m/pixel. The height at which this resolution is surpassed could be of significance in the design

of the DC. The DC should improve on the EIS’s NAC 0.5 m/pixel scale, in order to justify its

inclusion in a future penetrator mission. [17]

The EIS’s WAC and NAC are capable of taking colour images through pushbroom and framing

modes, using six broadband filters, with wavelengths ranging from 300 - 1050 nm. In Z. Turtle

(2015) the EIS bandpasses are presented, and are shown below in Table 3. [23]

Table 3 - EIS Filters

Name Wavelength

Range (nm)

Purpose

NUV 350-400 Surface colour. Plumes with Rayleigh scattering.

BLU 400-500 Surface colour. Rayleigh scattering with NUV.

GRN 500-600 Surface colour. Europa airglow during EIS eclipse or nightside

observation.

RED 650-780 Surface colour.

IR1 780-920 Surface colour. Provides a continuum for the H2O(s) band.

1µm 950-1050 Surface colour. Coarse grained H2O(s) band.

The EIS consists of two 4096 pixel × 2048 pixel (cross-track × along-track) CMOS detectors

that are radiation hardened. Radiation hardening significantly reduces the degradation of the

detector’s charge-transfer efficiency. The inclusion of a radiation hardened CMOS detector can

9

be justified in the design of a camera operating in the high radiation Jovian environment over an

extended period of time. However given that the DC is expected to operate for ≈ 4 − 5 minutes,

it may be the case that a non-radiation hardened instrument will suffice.

2.5 ESA’s JUICE Mission - JANUS

JUICE (JUpiter ICy moon Explorer) is an ESA L1 class mission, set to launch in 2022 to investigate

the characteristics and potential habitability of Ganymede, Callisto and Europa. The JANUS

(Jovis, Amorum ac Natorum Undique Scrutator) camera system is one of eleven instruments that

will form the payload of ESA’s JUICE mission. JANUS is a camera system with the ability

to perform high resolution imaging down to less than 10 m/pixel. Furthermore JANUS has a

1.72◦ × 1.29◦ FOV, 14 panchromatic broadband and narrow filters and a wavelength range from

0.36 µm to 1.1µm, along with the ability to take stereoscopic images. JANUS aims to answer

questions such as: How is the geological evolution of Europa related to its tectonic, impact and

cryovolcanic history? How old are specific geological units on Europa? And how do these findings

help explain the origin and evolution of the Jovian system? [18]

Figure 8: JANUS instrument design (left) and FWM (right), Schmitz et al. (2013) [18].

JANUS will be capable of providing 55% global coverage of Europa’s surface at ≤ 3 km/pixel,

35% global coverage at ≤ 1 km/pixel, 30% at ≤ 500 m/pixel, 1% at ≤ 100 m/pixel and 0.2% at

≈ 10 m/pixel. P. Palumbo1 et al. (2014) notes that all of the given resolutions are worse case

estimations and that a pointing strategy is under study in order to gain increased coverage at a

higher resolution. [23] JANUS consists of three main elements, which are:

1. An Optical Head. This includes a catadioptric telescope, a mounting structure, an aperture

cover, 14 panchromatic broadband and narrow filters, a filter wheel and a CMOS framing

detector.

2. Main Electronics. This includes modules for data handling, camera control, data com-

pression and power supply.

3. Proximity Electronics.

Elements of JANUS that are of particular relevance to the design of the DC are the 14 panchro-

matic broadband and narrow filters, the filter wheel and the CMOS framing detector. From N.

Schmitz et al. (2013) it appears that 9 of the JANUS filters relate directly to the study of the

surface of Ganymede, Callisto and Europa. From N. Schmitz et al. (2013) it is apparent that

central wavelengths and bandwidths had yet to be confirmed, although it is likely that these have

now been selected. The 9 relevant filters are displayed below in Table 4.

10

Table 4 - JANUS Filters [23]

Filter I.D. Wavelength

Range (nm)

Purpose

FPAN 400 - 900 Pan/mono-chromatic imaging.

FBLUE 410 - 490 Blue satellite colour imaging.

FGREEN 490 - 570 Green satellite colour imaging.

FRED 616 - 696 Red satellite colour imaging.

CTM Medium 930 - 950 Geological.

CTM Strong 740 - 760 Fe2+ detection.

Violet 370 - 450 Satellites surfaces UV slope.

NIR 1 870 - 950 Fe2+ detection.

NIR 2 925 - 1075 Fe2+ detection.

JANUS’s filter wheel, the Filter Wheel Module (FWM), positions the filters in front of the

CMOS framing detector through rotation, see Figure 8 (right). The design presented in N. Schmitz

et al. (2013) consists of a filter wheel with 14 filter windows and is driven by a step per motor

gear combination. Whilst the use of the FWM may be beneficial to JANUS, implementation of a

filter wheel may be unfavourable in the design of the DC.

Finally the CMOS detector that is intended to be used as JANUS’s detector is e2v’s CIS115.

The CIS115 has a 7µm pixel size and is based on e2v’s CIS107 detector. Furthermore a 1504×2000

pixel quadrant prototype has been developed, which is designed to allow for a 3008 × 4000 pixel

equivalent to be manufactured, P. Jorden et al (2014). [26]

3 Methods

This section presents the method of inquiry used in this research project relating to the theoretical

design of the DC. This section begins by presenting the scientific and instrument goals of the DC,

these include:

1. Obtaining high-resolution images of Europa’s surface at greater than 0.5 m/pixel

(i.e. greater than the best spatial resolution of the EIS’s NAC).

2. Obtaining panoramic colour and monochrome images of Europa’s surface.

3. Operating on a rapidly rotating penetrator or a PDS during descent from heights

of several km.

4. Aiding the penetrator’s astrobiological and geological in-situ measurements.

5. Operating at low illumination levels, possibly during periods close to Jupiter’s

aphelion.

3.1 Descent Camera Design Theory

This subsection presents a theoretical model of a DC, inspired by the CLEP mission concept. In the

CLEP mission concept the PDS is designed to perform spin-up at separation from the penetrator.

The spin axis of the PDS is intended to be parallel to the direction of freefall, which presents

an opportunity to design an unconventional DC that consists of a set of framelets as opposed to

a framing camera. [2] A framelet is defined as a narrow long strip of a detector. In a framing

camera an image is captured on an entire detector at a given time (this is the most common type

of camera). By creating a set of framelets parallel to the PDS’s spin axis allows for the DC to

implement a pushbroom or pushframe imaging method, similar to that used on LROC’s WAC. [15]

[16].

If the DC is attached to a PDS and the PDS is stationary at a time t = 0, at a height h(t) =

h(0) = H above Europa’s surface, then its initial velocity is u = 0 ms−1. Let gEuropa = 1.315 ms−2.

[3] From basic mechanics h(t) is found to be given by,

11

h(t) = H − gEuropat2

2.

Where t is the time after release. Moreover let the point A be the nadir, which is the point

directly beneath the PDS. [8] Let θ = 90◦ = π2 radians be the angle between the PDS’s rotation

axis and the plane in which the PDS rotates. Consider an nwidth × nlength element detector, that

with optics allows nlength to cumulatively subtend an angle θ. Let cj be the jth detector element

row. The angular span of each detector element row is given by,

α =θ

nlength.

Define the elevation angle ϕj as the angle at which detector element row cj looks at the surface

of Europa. [8] It should be noted that ϕj is governed by the orientation of detector and the choice

of optics. ϕj is measured from the rotation axis upwards to the plane in which the face of the PDS

is rotating.

Let the ground range be the horizontal distance from A to a point on the surface. [8] Therefore

by trigonometric arguments the ground range from the point that detector element row cj looks

at, at a time t is given by,

Rcj ,t =

(H − gEuropat

2

2

)tan(ϕj).

Where ϕj = jθnlength

. ϕj = ϕ90 = 90◦ = π2 radians corresponds to the case of the detector

element row that looks above Europa’s horizon. This angle is not considered in the model. Let

the angular velocity of the PDS be ω and define the dwell time td as the length of time a detector

observes a ground target, this is the time taken for the DC’s IFOV to sweep across a single ground

element area. [8] The IFOV is the field of view of a single detector element and for the detectors

presented in this report is typically on the order of a few µ radians. Furthermore let φ be the angle

swept out by a detector row consisting of nwidth detector elements, in the azimuthal direction in a

time tdnwidth. Additionally the area of a detector element row is given by area,

Acj = nwidth(Pixel Size)2.

Where the Pixel Size for a single detector element is typically ≈ a few µm.

The area of a ground element area on Europa’s surface at time t+1 is given by,

Ground Element Area = Acj ,cj+1,t+1 = Acj+1,t+1 −Acj ,t+1

Where,

Acj ,t+1 =Rcj ,t+1Rcj ,t sinφ

2,

Acj+1,t+1 =Rcj+1,t+1Rcj+1,t sinφ

2,

Acj ,cj+1,t+1 =sinφ

2(Rcj+1,t+1Rcj+1,t −Rcj ,t+1Rcj ,t).

The subscript notation used for Acj ,cj+1,t+1, cj , cj+1, t + 1, can be read as the area between

detector element rows cj and cj+1 at a time t+ 1.

If the ground element area for a particular elevation angle at a given time is represented by a

detector element row, then the spatial area resolution of DC can be approximated as,

Spacial Area Resolution =Acj ,cj+1,t+1

nwidth

Furthermore by calculating the first and second time differentials of a ground element area, for

12

Figure 9: Penetrator, PDS and DC model.

a given elevation angle, gives the time at which the ground element area changes at its maximum

rate. These first and second time differentials are given by,

d

dt

(Acj ,cj+1,t+1

)=Acj ,cj+1,t+1 −Acj ,cj+1,t

∆t,

d2

dt2(Acj ,cj+1,t+1

)=

1

∆t

(d

dt

(Acj ,cj+1,t+1

)− d

dt

(Acj ,cj+1,t

)).

3.2 Descent Camera Design Theory - Matlab Simulation Method

The theory presented in the previous subsection regarding the design of the DC was written as

a Matlab function. The function takes 5 user arguments, these include an initial velocity u, a

timestep ∆t, a freefall height H, a sampling number ns and a sampling angle φ and returns 3

vectors, R, Φmod and H. Expressing the spatial and angular elements of the DC as matrices

allows the DC’s motion, the ground element areas it observes and its spatial resolution to be

simulated. The Descent Camera Design Theory section was implemented as using the following

calculations.

H is a column vector with elements corresponding to the height of the PDS at time ti.

H =

h(t1)

h(t2)...

h(tm)

.

Ψ is a row vector corresponding to the elevation angles of the DC and is given by,

Ψ =(ϕ1 ϕ2 · · · ϕnlength

).

Φ is column vector that corresponds to the azimuth angle of the DC and is given by,

13

Φ =

φ1φ2...

φm

.

As the PDS completes multiple azimuthal rotations during its descent an element of Φ can

exceed 2π radians. This means applying mod(2π) to every element of Φ.

Φmod =

φ1mod(2π)

φ2mod(2π)...

φmmod(2π)

=

Φmod,1

Φmod,2

...

Φmod,m

.

Taking the tangent of all the Ψ elements gives the row vector T,

T =(tanϕ1 tanϕ2 · · · tanϕnlength

)=(T1 T2 · · · Tnlength

).

This then allows for calculation of all the ground ranges, Rcj ,t, and is given by,

R = H ·T =

h(t1)T1 h(t1)T2 · · · h(t1)Tnlength

h(t2)T1 h(t2)T2 · · · h(t2)Tnlength

......

. . ....

h(tm)T1 h(tm)T2 · · · h(tm)Tnlength

.

As all components of R can be determined an area matrix A can be calculated, which gives all

the ground element areas. A similar approach combine with using the Matlab function diff can be

used to calculate the first and second time differentials of A; namely A and A.

The model results presented in the next section considers 2 Ψ vectors, these include Ψ =

(1◦ 2◦ · · · 89◦), i.e. 1◦ to 89◦ in 1◦ increments and

Ψ =

[(π2

)( 1

1024

) (π2

)( 2

1024

)· · ·

(π2

)(1023

1024

)],

i.e. for an nwidth× 1024 element framelet. Note that ϕnlength6= 90◦ = π

2 radians, as calculation

of tanϕnlengthis required and tan π

2 is undefined. Furthermore φ = 2π1000 and the main release

height considered is H = h(0) = 35000 m. The model corresponding to the nwidth × 1024 element

framelet has nwidth = 14 and is shown in Figure 16. The model assumes that the optics are such

that the ground element area for a given elevation angle is divided by the 14 pixels.

3.3 Wavebands Method

The method used to select the optimal choice of wavebands to be observed by the DC was through

use of the literature review, research papers and matching the purpose of certain filters against the

scientific and mission objectives of the DC. The potential for the DC to perform astrobiological

science is a major influence in the choice of wavebands, presented in the next section.

3.4 Imaging Method

This subsection presents the method used to determine how the DC images Europa, through optical

physics, which ties in closely with the subtopic regarding the DC’s wavebands and the subsection

on the Descent Camera Design Theory. Whilst this subsection is similar to the subsection on the

Descent Camera Design Theory, it is beneficial in that it enables validation of the results by an

independent method.

14

Spatial resolution is defined as the smallest angular or linear separation between two objects

that is resolvable by a sensor. The focal length f of a detector, that is required to image an object

of radius R on the ground, is given by,

f

h(t)=rdR

= m.

Where rd is the detector array radius and m is the image magnification factor. As h(t) can

be determined and rd and R can be fixed, means that f can be determined as f = rdh(t)R . At the

diffraction limit of the detector the focal length f required to give an image of diameter D, for a

point target is, given by,

f =Dda2.44λ

.

Where da is the diameter of the camera’s aperture and λ is the wavelength of light that the

camera is observing. Therefore taking λ = λmin, for a fixed D and da, will give the maximum focal

length required by the camera.

For a diffraction limited camera the camera’s angular resolution ϕr is given by,

ϕr = 1.22λ

da.

And so by trigonometric arguments the ground resolution is given by,

∆x = 1.22λh(t)

da cosϕ.

Where ϕ is the elevation angle measured from the nadir vector, i.e. straight down from the DC.

The model results presented in the next section considers λmin = 410 nm, h(t) from the subsection

on the Descent Camera Design Theory. Also two values of D are considered, DJANUS = 10 m

and DEIS = 0.5 m, which correspond to the best spatial resolution of JANUS and the EIS’s

NAC respectively. The model will also use the expected camera apertures dJANUS = 0.1 m and

dEIS = 0.152 m.

3.5 Mass Method

This subsection presents the method used to determine the mass constraints on the DC. To obtain

an order of magnitude estimate of the DC’s mass, the masses for a set of DCs, used in past space

missions, was compiled, such as NASA’s Mars Science Laboratory (MSL) MARDI DC. The method

used to obtain a improvement on the DC’s mass, was through compiling a set of components that

should be included in the DC and then finding the masses of similar components from research

papers.

3.6 Power Method

This subsection considers the method used to obtain a power constraint on the DC. The author

used JPL’s HORIZONS Web-Interface, with the following parameters:

• Ephemeris Type: VECTORS.

• Target Body: Jupiter [599].

• Coordinate Origin: Solar System Barycenter (SSB).

This allowed the author to obtain Jupiter’s ephemeris from 15th March 2016 until 19th January

2035. By obtaining Jupiter’s ephemeris, Jupiter’s perihelion and aphelion distances could be found.

By obtaining these distances, the author was able to calculate the expected solar flux F in (Wm−2)

at perihelion and aphelion.

15

Assuming the Sun radiates isotropically as a blackbody, with a temperature T� = 5772 K, a

radius R� = 6.957× 108 m, with σ = 5.670367× 108 Wm−2K−4, then at a distance d in (A.U.),

F =4πR2

�σT4�

4πd2.

The above formula was written as a Matlab function. Results from this function are presented

in the next section.

3.7 Data Method

This subsection presents the method used to determine the data volume constraints and data rates

of the DC. Consider a single framelet on the DC, which is attached to a PDS, that consists of

nwidth × nlength detecting elements. Then in order to construct a single panoramic image, Ncap

framelet images must be captured, where Ncap is given by,

Ncap =2π

nwidthIFOVframelet.

Where IFOVframelet is the IFOV of a single detector element. Furthermore the number of

panoramic images Npan that can be generated will depend on the angular velocity of the PDS and

its impact time.

If the angular velocity ω = dφdt of the PDS is approximately constant during the camera’s

operation then the number of rotations Nrot is given by,

Nrot =ωtimpact

2π.

Which implies,

Npan = NcapNrot =ωtimpact

nwidthIFOVframelet.

Therefore if the data volume per image for a single framelet is given by Dimage, then the total

data volume Dt from the DC during its descent is given by,

Dt = DimageNcapNrotNf

Where Nf is the number of framelets in the DC. The model results presented in the next

section assumes ω = 2π rad s−1 and that nwidth = 14 and nlength = 1024 such that a single

framelet consists of 14×1024 detector element. Furthermore the model takes IFOVframelet = 7 µm

and Nf = 7.

3.8 Digital Terrain Model Method

This subsection presents the method used to obtain DTMs of Europa’s surface. Initially the project

implemented a .tiff global mosaic of Europa, derived from images taken by NASA’s Voyager 1,

Voyager 2 and Galileo spacecraft, to create a model of Europa’s surface. The mosaic was based on

the assumption that Europa is a sphere of radius 1562.09 km. [19]

The elements of the .tiff array take values ranging from 0 - 255, i.e. 8-bit grayscale. The

altitude of Europa’s surface is then set to correspond to these values, with 0 corresponding to a

global minimum and 255 corresponding to a global maximum. The surface height z is then scaled

to the global height variation of Europa ≈ 400 m [3] and the x and y coordinates are scaled to the

spatial resolution, which ranged from 20 km/pixel (gap fill) to 200 m/pixel [19].

For the DTM shown in Figures 22 and 23 a similar method was applied to a 1645× 1510 pixel

mosaic image called PIA01125.jpg, which covers an area of 365 km × 335 km of Europa’s surface,

at 2.9◦ S and 234.1◦ W.

16

4 Results

This section presents the results of the research project. The section begins with the results

from the Matlab simulations of the DC and then presents the results of calculations regarding its

constraints.

4.1 Descent Camera Design Results

This subsection presents the results obtained from the Matlab simulations of the DC, from the

methods section Descent Camera Design Theory.

Figure 10: A set of height curves (red) and velocity (blue) of the DC against freefall time.

Figure 10 shows 10 DC height-time curves in red and 10 DC velocity curves in blue. Where

the height h(t) is the distance of the DC from Europa’s surface and time t is the time after release

of the DC from stationary. As expected it can be seen that the greater the height at t = 0,

h(t) = h(0) = H, the later the DC impacts Europa. Figure 10 shows that after falling from

stationary from a height of 35000 m the DC impacts Europa’s surface at timpact = 230.72 s, this

is the intended h(0) for the CLEP penetrator and PDS. Furthermore for h(0) = 5000 m (i.e. the

lowest h(0) shown), timpact = 87.2041 s and for h(0) = 50000 m (i.e. the greatest h(0) shown),

timpact = 114.6733 s.

It can also be seen that the DC’s velocity curves overlap and that the DC’s velocity increases at

a constant rate and is greatest at timpact, with an impact velocity vimpact = 303.4 ms−1. Moreover

for h(0) = 5000 m, vimpact = 275.7637 ms−1 and for h(0) = 50000 m, vimpact = 362.6283 ms−1.

Plotting R and Φmod as a polar plot for Ψ = ( 44π180 ,

π4 ,

46π180 ) = (44◦, 45◦, 46◦) shows a trace of the

ground range for three adjacent detector element rows, this is shown in Figure 11. In Figure 11

the perspective to the left is a plan view and the perspective to the right is an oblique. Whilst not

shown in this visual, the plane of the trace is Europa’s surface. It is immediately apparent that an

interference pattern is present in the plot. It can be seen that Φmod is plotted in an anticlockwise

sense. Further investigation shows that the maximum ground range for the detector element row

with elevation angle 44◦ is R44,0 = 3.2638× 104 m, for 45◦ is R45,0 = 3.3799× 104 m and for 46◦

is R46,0 = 3.5000× 104 m.

17

Figure 11: Polar plot of the ground range from plan (left) and oblique (right) perspectives.

Whilst the interference pattern is an interesting phenomena, it obscures characteristics that

help explain how the ground range Rcj ,t changes with time and elevation angle. To obtain further

insight into the evolution of the ground range requires zooming in on the polar plot, this is presented

in Figure 12.

Figure 12: Zoomed polar plot edge (left) and centre (right).

Figure 12 (left) shows a zoomed in perspective of the edge of the polar plot. From Figure 12

(left) it can be seen that the greater the elevation angle the greater the maximum radial extent of

the Rcj ,t. Figure 12 (left) also shows that the ground range traces overlap. Figure 12 (right) shows

a zoomed in perspective of the centre of the polar plot. The centre of the polar plot is the impact

location of the DC. From Figure 12 (right) it can be seen that the ground traces spiral inwards in

an anticlockwise manner. Figure 12 (right) shows that the lower the elevation angle the small the

ground range is for a given time. This can be seen as the green ground range trace (44◦) is always

the inside trace, the red ground range trace (45◦) is always the central trace and the blue ground

18

range trace (46◦) is always the outside trace.

Figure 13: A set of ground element area curves for elevation angles from 1◦ to 10◦.

Plotting the ground element area A against time shows that, for a given time, for h(0) = 35000

m, the greater the elevation angle the greater the ground element area, see Figure 13. Figure 13

shows a set of ground element area curves against time, with the lowest curve corresponding to

ϕ1 = 1◦ the and the highest curve corresponding to ϕ10 = 10◦. Figure 13 shows that the ground

element area decreases as time increases and is 0 at timpact. Furthermore Figure 13 shows that

the initial ground element area at t = 0 for ϕ2 = 2◦ is Ac2,c3,1 = 1.1725× 103 m2 and similarly for

ϕ10 = 10◦ is Ac10,c11,1 = 2.0527× 104 m2.

Figure 14: A set of ground element area curves for elevation angles from 1◦ to 45◦.

Figure 14 shows the ground element area curves for elevation angles from 1◦ to 45◦ and similarly

Figure 15 shows the ground element area curves for elevation angles from 1◦ to 89◦. In Figure

13 it appears that the ground element area, at any fixed time, increases at a constant rate as the

elevation angle increases. However by comparing Figures 13 and 15 it can be seen that the ground

19

Figure 15: A set of ground element area curves for elevation angles from 1◦ to 89◦.

element area, at any fixed time, increases at an increasing rate as the elevation angle increases.

This is apparent in Figure 15, as it appears that every ground element area curve with an elevation

angle below ≈ 80◦ appears to be a single black line on the Time-after-release axis.

Figures 13, 14 and 15 also show that there is a point in time during the DC’s descent when

the ground element area is decreasing at a maximum rate. By calculating the first and second

time differentials of the ground element areas it is found that the maximum change of a ground

element area occurs at tmax(A) = 133.206 s and appears to be independent of the elevation angle.

Interestingly the following appears to hold for all elevation angles between 1◦ and 89◦ and for all

positive release heights,

Λ =tmax(A)

timpact= 0.5773 = Constant.

Figure 16: A model of 14 × 1024 element framelet showing the correspondence between detector element

row and ground element.

20

Where tmax(A) is the time at which the ground element areas are decreasing at their maximum

rate and timpact is the impact time.

By considering the ground element areas for a 14× 1024 element framelet involves calculating

the ground element areas for,

Ψ =

[(π2

)( 1

1024

) (π2

)( 2

1024

)· · ·

(π2

)(1023

1024

)].

The correspondence of a framelet detector element row to a ground element area curve is

illustrated in Figure 16. The blue line in Figure 16 corresponds to the best spatial resolution of

the EIS’s NAC. This method is similar to that of ray tracing and if time allowed this would be

attempted. From Figure 16 it can be seen that Row 1 corresponds to the first ground element area

curve, Row 2 corresponds to the second ground element area curve et cetera.

Figure 17: A set of ground element area curves for the first (i.e. lowest elevation angle) 128 rows of a

14 × 1024 element framelet.

Figure 17 shows the results for the 14 × 1024 detector element framelet, for the 128 lowest

elevation angle rows, which represents 18

thof the framelet’s length. As expected, it is apparent

that the ground element area curves are more tightly spaced than the 1◦ to 89◦ model. Figure 17

also shows the best spatial resolution of JANUS, represented by the horizontal red line at 100 m2

and the EIS’s NAC by the blue horizontal line 0.25 m2. It should be noted that the EIS’s NAC

line is difficult to see due to its proximity to the Time-after-release axis. In a similar manner

to Figure 11 it is also apparent that an interference pattern is present, especially close to the

Time-after-release axis near to timpact.

21

Figure 18: Results from a Matlab simulation of a 14× 1024 element framelet of the DC, showing the points

in time when detector element rows of the DC’s spatial resolution exceeds the best resolution of JANUS

and the EIS’s NAC. Release height 35000 m.

Figure 18 shows the set of times for which the 14 × 1024 framelet exceeds the best spatial

resolution of JANUS and the EIS’s NAC, for a release height of 35000 m. The y-axis represents

the detector element row, with the row number increasing with elevation angle. The graph shows

that as time increases the number of detector element rows that exceed the best spatial resolution

of JANUS and the EIS’s NAC increases at an exponential rate. The graph also shows that at

t = 0, 76 detector element rows already exceed the best spatial resolution of JANUS. On further

investigation of Figure 18, it can be seen that at timpact, 968 detector element rows beat the best

spatial resolution of JANUS and 613 detector element rows beat the best spatial resolution of the

EIS’s NAC. However the time at which the first detector element row achieves this is 141.9090 s

after the PDS’s release. Furthermore 582 of these detector element rows only achieve better than

this < 10 s before timpact.

22

Figure 19: Results from a Matlab simulation of a 14 × 1024 element framelet of the DC. Release height

5000 m as opposed to 35000 m.

Figure 19 shows the results for the 14 × 1024 detector element framelet, with a release height

of 5000 m instead of 35000 m. Figure 19 shows a significant change from Figure 18. On further

investigation it is found that at timpact, 995 detector element rows beat the best spatial resolution of

JANUS and 809 detector element rows beat the best spatial resolution of the EIS’s NAC. However

this time 9 detector element rows exceed the best spatial resolution of the EIS’s NAC at t = 0.

Furthermore 662 detector element row beat the best spatial resolution of JANUS at t = 0.

4.2 Wavebands Results

This subsection presents the results obtained from the methods section Wavebands Method and

addresses the ideal layout and filters for the DC. From the literature review and from additional

research papers, such as Greeley et al. (2009), the author suggests that the filters, shown in Table

5, be included in the design of the DC.

Figure 20: EIS (left)and JANUS (right) filter wavelength ranges.

23

Figure 21 shows the DC’s theoretical structure. The DC consists of a 1024× 1024 pixel CCD

(or CMOS) detector, with the 7 narrow long colour filters, shown in Table 5 forming a CFA, which

is intended to cover 7 framelets. It can also be seen from Figure 21 that the DC’s theoretical

structure is similar in design to the MRO’s MARCI instrument.

Figure 21: DC theoretical structure (left). MRO MARCI structure (right), Bell et al. (2009).

Table 5 - Descent Camera Filters

Filter I.D. Wavelength

Range (nm)

Purpose

α (Visible) 410 - 490 Blue satellite colour imaging.

β (Visible) 490 - 570 Green satellite colour imaging.

γ (Visible) 570 - 616 Yellow satellite colour imaging.

δ (Visible) 616 - 696 Red satellite colour imaging.

ε (Visible-NIR) - Visible-NIR continuum.

ε (NIR) 1000 - 2000 Hydrated sulfates and brines.

ζ (MIR) 5000 - 7000 Characterise and map organic chem-

istry and potential biosignatures. C-O,

C-C and C-N bonds.

4.3 Imaging Results

Given the time limitations of the project, results from the subsection Imaging Methods in the pre-

vious section are not presented. The next step would be to implement the theoretical model into a

Matlab simulation and compare the results against the results from the Descent Camera Design Results

subsection.

4.4 Mass Results

The mass allowance of ESA’s contribution to NASA’s Europa Mission was expected to be in the

region of 250 kg. This is no longer available. [1] The lack of a defined mass allowance makes it

challenging to implement a mass constraint on the DC. However the typical masses of DC’s used

in previous missions suggests that the DC should have a mass on the order ≈ 1 kg. It should be

noted that this is a rough estimate.

Table 6 - Descent Camera Masses [12] [13] [14]

Camera Mass (kg)

MPL (MARDI) < 0.5

MSL (MARDI) 0.6

Cassini Huygens (DISR) 7.4

LROC WAC 0.9

24

The table below lists the main components that the author suggests should be included in the

DC. If time allowed the next logical step would be to select an ideal set of components and measure

their masses, so as to obtain a more rigorous estimate of the DC’s total mass.

Table 7 - Descent Camera Components

DC Component Purpose

Visible Lens Focuses visible light onto the CCD or CMOS detector.

NIR Lens Focuses NIR light onto the CCD or CMOS detector.

MIR Lens Focuses MIR light onto the CCD or CMOS detector.

CCD/CMOS Detector Measures light in a high radiation environment.

CFA See Table 5.

Mounting Structure Provides infrastructure and support for the DC.

Prisms Allows visible, NIR and MIR lenses to image on the same

CMOS detector.

Baffle(s) Blocks light sources outside the FOV of the DC, such as

Jupiter, Ganymede, Callisto, Io and the Sun.

4.5 Power Results

This subsection presents results of the power constraints imposed on the DC.

Table 8 - Solar Flux at Jupiter’s Future Aphelions and Perihelions

Date Orbital

Phase

Distance

(A.U)

Solar

Flux

(Wm−2)

11/02/17 Aphelion 5.45233 45.7873

16/01/23 Perihelion 4.94222 55.7270

31/12/28 Aphelion 5.45301 45.7759

13/12/34 Perihelion 4.94701 55.6191

The expected arrival date of a mission to Europa is in the late 2020’s. From the above table

it can be seen that near to the expected arrival date of a mission to Europa, Jupiter is close to

aphelion at 5.45301 A.U.. At this distance the solar flux is approximately 17.9 % less than at

perihelion in 2023. The implications of the reduction in solar flux on the DC are briefly discussed

in the next section.

4.6 Data Results

Given the time limitations of the project, results from the subsection Data Method in the previous

section are not presented. The next step would be to implement the theoretical model into a

Matlab simulation, then compare and analyse the results against the downlink rates from CLEP.

25

4.7 Digital Terrain Model Results

This subsection presents the DTM results. Figure 22 shows a framelet image simulation combined

with a DTM corresponding to PIA01125.jpg. The figure shows a set of 14× 1024 pixel framelets,

which when combined, form a panoramic image of Europa’s surface. Ideally this is how images

captured by a single framelet would appear.

Figure 22: Framelet image simulation combined with a DTM.

Figure 23 shows a 14 × 1024 framelet simulation, combined with a DTM, corresponding to

PIA01125.jpg, that traces the ground range from 2 detector element rows, Row 768 and Row 512.

This was implemented using the Matlab function comet3. From Figure 23 it can be seen that the

maximum ground range from Row 768 is almost twice that of Row 512.

Figure 23: A 14 × 1024 framelet simulation combined with a DTM.

26

5 Discussion

In this section the author examines and analyses the results presented in the previous section. The

author aims to interpret and explain the results as well as relate the results to the literature review.

5.1 Descent Camera Design Discussion

The PDS’s impact time timpact = 230.72 s and its impact velocity vimpact = 304.4 ms−1, presented

previously in the results section, closely agree with ESA’s CDF Study Report (CLEO/P). In ESA’s

report the PDS’s timpact = 231 s and its vimpact = 300 ms−1. The author believes that the slight

difference in these results is simply due to rounding. Confirming this basic finding then allowed

the author to further develop the DC model.

The author notes that this motion is modeled on the assumption that the PDS passes through

a vacuum from its release height until impact. If the penetrator is to investigate landing sites

of significant scientific interest, there is the possibility that the penetrator and PDS could pass

through a H2O plume, similar to plumes observed on Enceladus by NASA’s Cassini spacecraft.

Nimmo et al. (2007) has found that analogous plumes would be ≈ 70 km high when scaled to

Europa’s gravity, which is greater than the maximum descent heights that the DC and CLEP were

modeled with. Passing through such a plume could introduce drag forces on the penetrator or

PDS during descent. This would mean that the motion modeled in this report would be invalid.

A critical point in the DC’s descent is the point in time during its descent when its imaging res-

olution exceeds 10 m/pixel, the best resolution of the JANUS, and 0.5 m/pixel, the best resolution

of the EIS’s NAC.

The author now asks:

• Why are these times important?

By knowing the times when the DC’s spacial resolution exceeds that of JANUS and the EIS’s

NAC and how the set of times change, given different parameter’s, means that an optimal imaging

strategy can be found. This means that the DC obtains results that are more relevant to its

scientific goals.

The author now asks:

• Why does the ground element area for a given time change at an increasing rate

as the elevation angle increases?

The answer becomes apparent by manipulation of Acj ,cj+1,t+1.

Acj ,cj+1,t+1 = h(t) tan(ϕj+1)h(t+ 1) tan(ϕj+1)− h(t) tan(ϕj)h(t+ 1) tan(ϕj)

Acj ,cj+1,t+1 = h(t)h(t+ 1)[tan(ϕj+1)2 − tan(ϕj)

2]

Therefore Acj ,cj+1,t+1 can be written as two functions, each with independent variables.

g(t) = h(t)h(t+ 1)

f(j) = tan(ϕj+1)2 − tan(ϕj)2

Therefore,

Acj ,cj+1,t+1 = g(t)f(j).

So for a given time, g(t) is a fixed value. However f(j) is independent of t. And as tanϕj is

an exponentially increasing function for 0 < ϕj <π2 , f(j) exponentially increases as ϕj increases.

Therefore the ground element area for a given time increases at an increasing rate as the elevation

angle increases.

27

The author now asks:

• Why is the ground element area decreasing at an increasing rate before tmax(A)

and decreasing at a decreasing rate after tmax(A)?.

The answer becomes apparent by manipulation and differentiation of Acj ,cj+1,t+1. Given time

limitations the author is unable to present a rigorous explanation. However the explanation involves

fixing ϕj , calculating ddt

(Acj ,cj+1,t+1

)and d2

dt2

(Acj ,cj+1,t+1

)and setting d2

dt2

(Acj ,cj+1,t+1

)= 0.

The author now offers an explanation as to:

• Why the length of time that a fixed detector element row, exceeds the best spatial

resolution of JANUS and the EIS’s NAC, increases as the height at which freefall

begins, decreases?

Consider a penetrator or PDS that is released from a height h1,t = h1,0. At some time t1,

where 0 < t1 < timpact, the penetrator or PDS is at a height H and is traveling at velocity v1,t1having accelerated from v1,0 = 0 ms−1 in a time t1, at a constant rate of gEuropa = 1.315 ms−2.

If the height at which the penetrator or PDS decreases to h2,t = h2,0 then the time at which the

penetrator or PDS reaches a height H t2, is such that t2 < t1. Therefore given that the penetrator

or PDS accelerate at a constant rate implies that v1,t1 > v2,t2 . Therefore the DC travels faster

at H (if released from a greater height) and so spends less time observing at better resolutions.

This suggests that it is preferable to release the penetrator or PDS from a lower altitude, however

technically this may be more challenging. Furthermore it would mean that the DC would operate

for a shorter length of time as the timpact is less when released from a lower height. This would also

mean that the penetrator and the PDS would impact the surface of Europa at a lower velocity.

This would mean that the penetrator would not penetrate as deep into Europa, which could reduce

the penetrator’s astrobiological scientific return.

5.2 Wavebands and Imaging Discussion

This subsection discusses the waveband and imaging results obtained for the DC. A significant

driver in the choice of wavebands is the emission spectra of Europa’s surface. An important

waveband to consider in the design of the DC was initially the IR region from 0.7 µm to 3µm. [8]

This waveband consists of the NIR, 0.7 µm to 3.0 µm and the MIR from 3.0 µm to 50 µm and is

frequently used in remote sensing. [8]

The author is surprised to find that whilst many past and future Europan dedicated camera’s

are designed to incorporate cameras capable of UV, visible and NIR imaging, very few appear to

incorporate the MIR into their design. There may be justifiable reasons for this that the author

is unaware of. This could be due to the fact that signal intensity in the MIR is less than the UV,

visible and NIR signal intensities. However from Greeley et al. (2009) it appears that obtaining

images at wavelengths from 5µm to 7µm offers the ability to perform important astrobiological

science. Obtaining images and spectra at wavelengths in the 5µm to 7µm region would allow

mapping and characterisation of organic chemistry on Europa’s surface, such as C-O, C-N and

C-C bonds, all of which are associated with life.

5.3 Mass, Power and Data Discussion

This subsection discusses the mass, power and data results obtained for the DC. Given the time

constraints on this project limited progress has been made with regards to the mass, power and

data.

5.4 Digital Terrain Model Discussion

This subsection discusses the digital terrain model results obtained for the DC. The primary

purpose of creating was the DTM was to become familiar in creating models of surfaces in Matlab

and to apply techniques to manipulate these surfaces.

28

Whilst the implementation of shape from shading (SFS) algorithms to several greyscale images

can be used to put constraints on the surface altitude, a single greyscale image does not. [20]

Only single grayscale images were used to create the DTMs in this project. Therefore the DTMs

presented are not an accurate representation of the real terrain, so their scientific use is limited.

Although they are not accurate, they appear very realistic and so are useful in simulating the DC’s

theoretical FOV, especially when viewed from an oblique perspective, as is shown in Figure 22.

6 Conclusions

This section presents a set of conclusions from the project results and discusses the extent to which

the main results and discussion meet the aims and objectives of the project, and finally discusses

possible future research that could follow on from the project. The main conclusions of the research

project are as follows:

• A DC based on the theoretical model presented is potentially of scientific benefit for a Europa

penetrator mission, as it would be capable of obtaining colour panoramic images at a greater

spatial resolution than JANUS (≈ 10 m/pixel) and the EIS’s NAC (0.5 m/pixel).

• The length of time that a fixed detector element row exceeds the best spacial resolution of

JANUS and the EIS’s NAC increases as the height at which freefall begins decreases.

• The DC may need to operate in illumination levels that are up to 17.8% less than at Jupiter’s

aphelion.

• Inclusion of MIR imaging in the DC may offer significant astrobiological scientific return.

The author highlights that the results from the Matlab simulations of the DC’s operation should

be taken with caution.

6.1 Future Research and Improvements

This subsection presents future research and improvements that, given more time, the author would

like to investigate.

• The Matlab models used φ = 2π1000 and a time step φ = 2π

1000 , giving a sampling rate dφdt ≈

0.0063 radians s−1. Whilst fixing these parameters enabled investigation and development

of the DC it is likely that they will significantly influence the results.

• The author has included visible, NIR and MIR lenses in the DC design, based on the design of

other cameras, with little justification. Given more time the author would like to investigate

whether three lenses would actually be required. Reducing the number of lenses would reduce

the DC’s mass, but could potentially reduce the quality of the images obtained.

• Whilst a CMOS detector appears favourable for the DC, if more time was allowed the author

would like to justify more rigorously why a CMOS detector is the most suitable type of

imaging detector.

• Given more time the author would increase the axis number font size, as it can be challenging

to read certain graphs.

• The project aimed to create a 3D map of Europa’s surface in Term 3 using DTM files from

Dr. Paul Schenk, given time constraints the author was unable to obtain these files, however

if these files can be obtained they provide an excellent opportunity for further research.

29

7 Acknowledgments

I wish to acknowledge the guidance of my project

supervisor Dr. Geraint Jones.

8 Bibliography

All simulations, illustrations, figures (except Fig-

ures 1-8) are by the author.

References

[1] Clipper ESA Orbiter (CLEO) and

Clipper ESA Penetrator (CLEP),

http://sci.esa.int/trs/56474-clipper-esa-

orbiter-cleo-and-clipper-esa-penetrator-

clep/, ESA, 18 September 2015.[2] MSc Penetrators Project List, Geraint

Jones, MSSL - UCL, 2015.[3] Planetary Science Second Edition, Imke de

Pater and Jack J.Lissauer, Cambridge Uni-

versity Press, 2010.[4] Figure 1, Europa Chaos Terrain, www.

nasa.gov/multimedia/imagegallery/

image_feature_1339.html, NASA,

28/04/15.[5] Figure 2, PIA01669: Model of

Europa’s Subsurface Structure,

http://photojournal.jpl.nasa.

gov/catalog/PIA01669, NASA JPL,

19/01/99.[6] CFD Study Report CLEO/P Assessment

of a Europa Penetrator Mission as Part of

NASA Clipper Mission, Concurrent Design

Facility, European Space Agency, April

2015.[7] Figure 3, CFD Study Report CLEO/P As-

sessment of a Europa Penetrator Mission

as Part of NASA Clipper Mission, Con-

current Design Facility, European Space

Agency, April 2015.[8] Spacecraft Sensors, Mohamed M Abid,

Wiley, 2005.[9] Charge-coupled device spectra of the

Galilean satellites: Molecular oxygen on

Ganymede W.M. Calvin et al, Journal of

Geophysical Research, Sept 1995.[10] Solar System Ices, B.Schmitt, C.DeBergh

and M.Festoan, ASSL, 1996.[11] A multi-instrument spectral view of Eu-

ropa from Galileo, J.C. Granahan, et

al, Lunar and Planetary Science XXVIII,

1997.

[12] The MAST cameras and the Mars Descent

Imager (MARDI) for the 2009, Mars Sci-

ence Laboratory, Lunar and Planetary Sci-

ence XXXVI, M. C. Malin et al, 2005.[13] Mars Descent Imager (MARDI) on the

Mars Polar Lander, Journal of Geophysi-

cal Research, Malin et al, August 2001.[14] The Descent Imager/Spectral Radiometer

(DISR) Aboard Huygens, M. G. Tomasko

et al, 2002.[15] Lunar Reconnaissance Orbiter Camera

(LROC) Instrument Overview, M.S.

Robinson et al, Springer Science+Business

Media B.V., 2010.[16] Spacecraft Systems Engineering Fourth

Edition, Peter Fortescue, Graham Swinerd

and John Stark, Wiley, 2011.[17] The Europa Imaging System (EIS): High-

resolution imaging and topography to in-

vestigate Europa’s geology, ice shell, and

potential for current activity, 47th Lunar

and Planetary Science Conference, E. P.

Turtle et al. 2016.[18] JANUS on JUICE: A Camera to Investi-

gate Ganymede, Europa, Callisto and the

Jovian System, N. Schmitz et al.[19] Figure 4, Europa Voyager and Galileo

SSI Global Mosaic 500m, http:

//astrogeology.usgs.gov/search/

details/Europa/Voyager-Galileo/

Europa_Voyager_GalileoSSI_global_

mosaic_500m/cub, U.S Geological Survey,

January 2010.[20] Shape From Shading: A Method for Ob-

taining the Shape of a Smooth Opaque

Object From One View, Berthold K. P.

Horn, MIT Artificial Intelligence Labora-

tory, 1970.[21] The internal structure models of Europa,

JIN Sheng and JI JiangHui, Science China

Physics, Mechanics and Astronomy, Jan-

uary 2012.[22] Figure 1, Juno Closes in on Jupiter,

http://www.nasa.gov/image-feature/

pia20706/juno-closes-in-on-jupiter,

NASA, July 2016.[23] JANUS: The Visible Camera Onboard The

ESA JUICE Mission To The Jovian Sys-

tem, 45th Lunar and Planetary Science

Conference, 2014.[24] High-Resolution, 3-D Insight into Europa’s

Ice Shell and Potential for Current Activ-

ity, John Hopkins Applied Physics Labo-

ratory, 2015.

30

[25] Future Exploration of Europa, Europa,

Robert T. Pappalardo, William B. McKin-

non and Krishan Khurana, Greeley et al.,

2009.[26] e2v new CCD and CMOS technology de-

velopments for astronomical sensors, e2v,

P.R Jorden et al., 2014.

31