Revisiting the Phoenix TECP data: Implications for
regolith control of near-surface humidity on Mars
Edgard G. Rivera-Valentina , Vincent F. Chevrierb
a
Arecibo Observatory, Universities Space Research Association, HC 3 Box 53995,
Arecibo, PR 00612, USA
b
Arkansas Center for Space and Planetary Sciences, University of Arkansas,
Fayetteville, AR 72701, USA
Abstract
We analyze the recalibrated in-situ humidity measurements by the Phoenix
lander, which landed at 68.2◦ N, 234.3◦ E and operated from Ls 78◦ through
Ls 148◦ , ∼ 152 sols. Vapor pressures demonstrate significant day-night variations with values ranging from 0.005 Pa to 0.37 Pa, an order of magnitude
lower than previously reported, and evening derived enthalpies less than that
for a purely ice deposition-driven process, suggesting other water vapor sinks
may contribute to the near-surface humidity at the martian polar regions.
Keywords: Mars, Mars, atmosphere, Meteorology
1
1. Introduction
2
One of the primary goals of the Phoenix lander mission (PHX), which
3
landed in the martian sub-polar terrain, was to investigate water vapor trans-
4
port processes between the surface and atmosphere (Smith et al., 2008; Zent
5
et al., 2010). Operations lasted from Ls 78◦ through Ls 148◦ and so PHX was
6
able to characterize ∼ 80% of the northern summer. Phoenix provided the
7
first in-situ humidity measurements via its thermal and electrical conductivPreprint submitted to Icarnus
March 2, 2015
8
ity probe (TECP) (Zent et al., 2009). Initial diurnal water vapor variations
9
demonstrated a nearly Gaussian shape with a characteristic plateau during
10
the day and an amplitude not described by the purely atmospheric Global
11
Circulation Models (Chevrier et al., 2009). Additionally, water vapor pres-
12
sure demonstrated a decrease in the evening before relative humidity reached
13
RH = 100%, suggesting a water vapor sink other than condensation may be
14
controlling the near-surface water cycle (Zent et al., 2010). Several processes
15
have been suggested, including ice-atmosphere exchange through diffusion,
16
adsorption onto regolith grains (Zent et al., 2010), and hydration changes of
17
salts (Chevrier et al., 2009).
18
Previous analysis inferred a water vapor variation of 0.02 Pa ≤ PH2 O ≤ 2 Pa
19
(Smith et al., 2009; Zent et al., 2010). A new calibration function, though,
20
was provided later (Zent et al., 2012; Zent, 2014) that significantly reduced
21
the inferred PH2 O . Here, we present analysis of the recalibrated Phoenix
22
TECP data in order to elucidate the water cycle at the landing site. We find
23
local water vapor pressure variations between 0.005 Pa and 0.37 Pa, an or-
24
der of magnitude smaller than previously reported. By coupling in-situ data
25
with our thermal model (Chevrier and Rivera-Valentin, 2012; Nuding et al.,
26
2014), we derive enthalpy values for the atmosphere-surface exchange. We
27
find that the evening enthalpy is ∼ 13 kJ mol−1 , suggesting another water
28
vapor sink may be active at the Phoenix landing site.
29
2. Data Analysis
30
The TECP measured relative humidity with a capacitance-based relative
31
humidity sensor, which was located above the TECP needles, and tempera2
32
ture (T ) via a Type E thermocouple (Smith et al., 2008; Zent et al., 2009).
33
Temperature was also measured near the humidity sensor using the electron-
34
ics board sensor. The TECP conducted both atmospheric and in-soil mea-
35
surements, where air measurements were obtained between ∼ 0.14 m and
36
2.3 m (Zent et al., 2010). Measurement duration lasted between ∼ 13 min
37
and 11 hrs and data was nominally taken in ∼ 1 s intervals.
38
Here we analyze the recalibrated data. The previous reported calibra-
39
tion function provided unrealistic RH values due to the use of laboratory
40
calibration that included data acquired up to ∼ 300 K (Zent et al., 2012).
41
The revised calibration function now includes flight data taken when the at-
42
mosphere was saturated (Zent et al., 2012; Zent, 2014). Extracted values
43
from the NASA Planetary Data System database were sol, type of measure-
44
ment, time, temperature, relative humidity, and water vapor pressure, which
45
was inferred using the measured RH and the board temperature. In soil-
46
measurements are integrated over a length of ∼ 15 mm, which is the size
47
of the TECP needles, and therefore are affected by the steep temperature
48
gradient caused by the low regolith thermal conductivity (Zent et al., 2010).
49
Additionally, because of the location of the humidity sensor, during some
50
in-soil T measurements, RH was measured in-air. The bulk of the TECP
51
measurements, though, were made in-air (∼ 80%). Thus, we limit our anal-
52
ysis to in-air measurements. Data is binned and averaged over 1 hr intervals
53
to reduce instrument noise. Reported error is based on data variation.
54
The diurnal vapor pressure (Figure 1a) varies between 0.005 Pa ≤ PH2 O ≤
55
0.37 Pa. This is in contrast to the previously reported range of 0.02 Pa ≤
56
PH2 O ≤ 2 Pa (Smith et al., 2009; Zent et al., 2010). The average vapor
3
57
pressure during the day is 0.18 ± 0.03 Pa (∼ 15 pr-µm), which is an order of
58
magnitude lower than the previous value of 1.8 Pa (Zent et al., 2010), and
59
relative humidity varied between 0.06% - 7.2% with an average of 0.6%. The
60
diurnal curve is characterized by a plateau during the day, though not as
61
pronounced as was previously seen (Chevrier et al., 2009; Zent et al., 2010).
62
A significant drop in PH2 O occurs between midnight and local dawn when
63
RH reaches 100% and so is attributed to condensation, which was observed
64
at the landing site (Smith et al., 2009; Cull et al., 2010a). Condensed ice
65
could reach a thickness of a few mm and would quickly sublimate away in a
66
few minutes during sunrise. Frost formation was observed by Phoenix during
67
nighttime on sol 80 (Ls = 113◦ ), which is when RH ≈ 100% in Figure 1b.
68
The drop in vapor pressure after sunset, though, which is not as drastic,
69
occurs during a time when condensation was not observed. In fact, relative
70
humidity after sunset and before midnight does not reach 100% (Figure 1b),
71
so the drop in vapor pressure during this time may not be caused by conden-
72
sation. Additionally, the drop in PH2 O primarily occurs during the beginning
73
and end of the mission. For 90◦ . Ls . 130◦ , water vapor generally in-
74
creases in the evening. Maximum seasonal temperatures will occur around
75
Ls 90◦ and so increases in evening PH2 O are occurring during high tempera-
76
tures. This may suggest an Ls dependent process that is a sink during some
77
parts of the year and a source during the warmest part of the season, such
78
as ice-atmosphere exchange through diffusion, or a shadowing effect (Hecht,
79
2010).
4
80
3. Interpretation
81
The recalibrated TECP data indicates vapor pressure values are 10 times
82
smaller than previously reported. Maximum daytime vapor pressures are
83
still smaller than predicted by boundary-layer simulations, which suggest
84
maximal values of ∼ 0.7 Pa (Savijärvi and Määttänen, 2010). This may
85
imply an active regolith sink during the daytime hours. On the other hand,
86
the simulations may not be fully accounting for sublimation/condensation of
87
water ice (Savijärvi and Määttänen, 2010).
88
In order to identify the active mechanisms at the Phoenix landing site,
89
we use the PH2 O and T to infer the thermodynamics of ongoing processes. In
90
Figure 2, we construct a vapor pressure curve using all of the air measure-
91
ments, plotting the inverse temperature versus the log of the water vapor
92
pressure. The slope of these lines are indicative of enthalpy. We find there
93
is a significant inflection point near 220 K. For T > 220 K, the curve shows
94
a positive and/or neutral slope; this primarily occurs during the day. For
95
T < 220 K, there is a clear negative slope, which occurs primarily during
96
early morning and evening hours. Additionally, the morning and evening
97
hours have distinct slopes and therefore different enthalpies.
98
Using the earlier data set, Zent et al. (2010) found ∆H ∼ 34 kJ mol−1 dur-
99
ing the evening, far less than the enthalpy of ice-deposition (50.9 kJ mol−1 ).
100
Condensation will occur on the coldest surface, which during nighttime is the
101
surface of the soil and during the day is the surface of the ice (Hecht, 2010).
102
Because the dataset only includes in-air measurements, we join the Phoenix
103
in-situ data to our thermal model (Chevrier and Rivera-Valentin, 2012; Nud-
104
ing et al., 2014) in order to provide further quantitative constraints. Surface
5
105
and subsurface temperatures are simulated via two models: (a) assuming a
106
two-layer system with an ice table depth of 5 cm (Mellon et al., 2009), and
107
soil thermal inertia of Γ ∼ 300 J m−2 K−1
108
surements (Zent et al., 2010), and (b) assuming a homogeneous subsurface
109
with Γ ∼ 150 J m−2 K−1
110
(Mellon et al., 2008). The thermal model simulates heat transfer to a depth
111
of 4 m, well below the seasonal skin-depth, with finite element thickness of
112
0.005 m and time step of 10 s.
−1/2
−1/2
, consistent with in-situ mea-
, consistent with remote sensing observations
113
In Figure 3, we show the PH2 O − T diagram derived using our simulated
114
temperatures from both models. Here we assume a well-mixed lower atmo-
115
sphere such that PH2 O at the height measured by Phoenix is the same at the
116
surface. Enthalpy is derived from a least squares fit with 90% confidence. Be-
117
tween midnight and local dawn, our derived enthalpy of (49 ± 17) kJ mol−1
118
and (42 ± 18) kJ mol−1 , for models (a) and (b) respectively, corresponds well
119
with condensation. During the evening hours when a drop in PH2 O occurs,
120
we derive an enthalpy of (14 ± 5) kJ mol−1 and (11 ± 4) kJ mol−1 , while the
121
enthalpy is (−9 ± 3) kJ mol−1 and (−7 ± 2) kJ mol−1 when PH2 O increases
122
during the evening hour. Derived enthalpies between models are statistically
123
indistinguishable; however, enthalpy values are distinguishable between time
124
periods. Thus for T . 200 K (i.e., early morning), ice deposition is the
125
dominant water vapor sink at Phoenix. Results suggest that between Ls 78◦
126
and 90◦ and after Ls 130◦ there may be a water vapor sink other than con-
127
densation active in the evening hours (i.e., 200 . T . 230 K) at the martian
128
polar regions. On the other hand, for 90◦ < Ls < 130◦ water vapor increases
129
during the evening hours and a weakly negative enthalpy is inferred.
6
130
4. Discussion and Conclusions
131
We analyzed the recalibrated Phoenix TECP humidity data. Water va-
132
por pressure ranges from 0.005 to 0.37 Pa, nearly an order of magnitude
133
smaller than previously reported (Smith et al., 2009; Zent et al., 2010). The
134
diurnal humidity curve is characterized by a nearly-steady water vapor pres-
135
sure during the day, with an average of 0.18 Pa, that decreases on either
136
side (Chevrier et al., 2009; Zent et al., 2010). The significant drop in wa-
137
ter vapor occurring in the predawn hours can be attributed to condensation
138
(Zent et al., 2010); however, values after 1800 LST decrease before relative
139
humidity reaches 100%. An evening drop in water vapor pressure primarily
140
occurs early and late in the mission; the mid-mission is characterized by an
141
increase in water vapor. This may suggest a seasonally dependent process.
142
Here we used in-situ water vapor pressure data and simulated temper-
143
atures to derive the enthalpy during the evening and early morning hours.
144
Our results suggest ∆H ∼ 45 kJ mol−1 during the early morning, which is
145
comparable to that of ice deposition (50.9 kJ mol−1 ). This agrees well with
146
observations of condensation during these hours (Smith et al., 2009; Cull
147
et al., 2010a). On the other hand, we find the enthalpy during the evening
148
hours to be quite low and to change signs with ∆H ∼ −8 kJ mol−1 during
149
the warmest days and ∆H ∼ 13 kJ mol−1 early and late in the mission.
150
Adsorption is a possible water vapor sink, but onto most mineral grains
151
it requires ∆H ∼ 80 kJ mol−1 (Zent et al., 2001; Chevrier et al., 2008;
152
Beck et al., 2010); however, this is only for the first monolayer. Smaller val-
153
ues occur for multilayer adsorbates. Indeed, the cohesive nature of the soil
154
surrounding Phoenix may suggest adsorption is an active water vapor sink
7
155
(Arvidson et al., 2009). Hydration of salts is potentially excluded because
156
PHX relevant salts generally require ∆H > 30 kJ mol−1 (Besley and Bottom-
157
ley, 1969). Deliquescence of perchlorate salts, which were detected by PHX
158
(Hecht et al., 2009), is suggested experimentally as a possible process at the
159
landing site, occurring primarily during evening and predawn hours (Gough
160
et al., 2011; Nuding et al., 2014). In fact, liquid formation is suggested based
161
on the distribution of salts in the regolith (Cull et al., 2010b) and measured
162
dielectric signatures (Stillman and Grimm, 2011). Because the measured
163
water vapor is less than suggested by boundary layer simulations (Savijärvi
164
and Määttänen, 2010) and inferred enthalpies are low, near-surface humidity
165
at the Phoenix landing site may be at least partially controlled by processes
166
other than condensation/sublimation during the late evening hours.
167
5. Acknowledgements
168
V. F. Chevrier acknowledges support from NASA MFRP NNX13AG67G
169
and E. G. Rivera-Valentin partial support from NASA NEOO NNX13AQ46G.
170
We thank two anonymous reviewers who helped enhance this manuscript.
171
6. References
172
Arvidson, R. E., Bonitz, R. G., Robinson, M. L., Carsten, J. L., Volpe, R. A.,
173
Trebi-Ollennu, A., Mellon, M. T., Chu, P. C., Davis, K. R., Wilson, J. J.,
174
Shaw, A. S., Greenberger, R. N., Siebach, K. L., Stein, T. C., Cull, S. C.,
175
Goetz, W., Morris, R. V., Ming, D. W., Keller, H. U., Lemmon, M. T.,
176
Sizemore, H. G., Mehta, M., 2009. Results from the Mars Phoenix Lander
177
Robotic Arm experiment. Journal of Geophysical Research 114, E00E02.
8
178
Beck, P., Pommerol, A., Schmitt, B., Brissaud, O., 2010. Kinetics of water
179
adsorption on minerals and the breathing of the Martian regolith. Journal
180
of Geophysical Research 115, E10011.
181
182
Besley, L. M., Bottomley, G. A., 1969. The water vapour equilibria over
magnesium perchlorate hydrates. J. Chem Thermodyn. 1, 13–19.
183
Chevrier, V., Ostrowski, D. R., Sears, D. W. G., 2008. Experimental study
184
of the sublimation of ice through an unconsolidated clay layer: Implica-
185
tions for the stability of ice on Mars and the possible diurnal variations in
186
atmospheric water. Icarus 196, 459.
187
Chevrier, V. F., Hanley, J., Altheide, T. S., 2009. Stability of perchlorate
188
hydrates and their liquid solutions at the Phoenix landing site, Mars. Geo-
189
physical Research Letters 36 (10), L10202.
190
Chevrier, V. F., Rivera-Valentin, E. G., 2012. Formation of recurring slope
191
lineae by liquid brines on present-day Mars. Geophysical Research Letters
192
39 (21), L21202.
193
Cull, S., Arvidson, R. E., Mellon, M., Wiseman, S., Clark, R., Titus, T.,
194
Morris, R. V., Mcguire, P., 2010a. Seasonal H2 O and CO2 ice cycles at
195
the Mars Phoenix landing site: 1. Prelanding CRISM and HiRISE obser-
196
vations. Journal of Geophysical Research 115, E00D16.
197
Cull, S. C., Arvidson, R. E., Catalano, J. G., Ming, D. W., Morris, R. V.,
198
Mellon, M. T., Lemmon, M., 2010b. Concentrated perchlorate at the Mars
199
Phoenix landing site: Evidence for thin film liquid water on Mars. Geo-
200
physical Research Letters 37 (L22203).
9
201
Gough, R. V., Chevrier, V. F., Baustian, K. J., Wise, M. E., Tolbert, M. A.,
202
2011. Laboratory studies of perchlorate phase transitions: Support for
203
metastable aqueous perchlorate solutions on Mars. Earth and Planetary
204
Science Letters 312, 371–377.
205
Hecht, M. H., 2010. Ice at the Phoenix landing site part V: The equilib-
206
rium strikes back. 41st Lunar and Planetary Science Conference Abstract
207
#1481.
208
Hecht, M. H., Kounaves, S. P., Quinn, R. C., West, S. J., Young, S. M. M.,
209
Ming, D. W., Catling, D. C., Clark, B. C., Boynton, W. V., Hoffman, J.,
210
Deflores, L. P., Gospodinova, K., Kapit, J., Smith, P. H., 2009. Detection
211
of Perchlorate and the Soluble Chemistry of Martian Soil at the Phoenix
212
Lander Site. Science 325, 64–67.
213
Mellon, M., Arvidson, R., Sizemore, H., Searls, M., Blaney, D., Cull, S.,
214
Hecht, M., Heet, T., Keller, H., Lemmon, M., 2009. Ground ice at the
215
Phoenix Landing Site: Stability state and origin. Journal of Geophysical
216
Research 114 (E1).
217
Mellon, M. T., Boynton, W. V., Feldman, W. C., Arvidson, R. E., Titus,
218
T. N., Bandfield, J. L., Putzig, N. E., Sizemore, H. G., 2008. A prelanding
219
assessment of the ice table depth and ground ice characteristics in Martian
220
permafrost at the Phoenix landing site. Journal of Geophyical Research
221
113, 1–14.
222
Nuding, D. L., Rivera-Valentin, E. G., Davis, R. D., Gough, R. V., Chevrier,
223
V. F., Tolbert, M. A., 2014. Deliquescence and efflorescence of Calcium
10
224
Perchlorate: An investigation of stable aquoeous solutions relevant to
225
Mars. Icarus 243, 420–428.
226
Savijärvi, H., Määttänen, A., 2010. Boundary-layer simulations for the Mars
227
Phoenix lander site. Quarterly Journal of the Royal Meteorological Society
228
136 (651), 1497–1505.
229
Smith, P. H., Tamppari, L., Arvidson, R. E., Bass, D., Blaney, D., Boynton,
230
W., Carswell, A., Catling, D., Clark, B., Duck, T., Dejong, E., Fisher, D.,
231
Goetz, W., Gunnlaugsson, P., Hecht, M., Hipkin, V., Hoffman, J., Hviid,
232
S., Keller, H., Kounaves, S., Lange, C. F., Lemmon, M., Madsen, M., Ma-
233
lin, M., Markiewicz, W., Marshall, J., Mckay, C., Mellon, M., Michelangeli,
234
D., Ming, D., Morris, R., Renno, N., Pike, W. T., Staufer, U., Stoker, C.,
235
Taylor, P., Whiteway, J., Young, S., Zent, A., 2008. Introduction to special
236
section on the Phoenix Mission: Landing Site Characterization Experi-
237
ments, Mission Overviews, and Expected Science. Journal of Geophysical
238
Research 113.
239
Smith, P. H., Tamppari, L. K., Arvidson, R. E., Bass, D., Blaney, D., Boyn-
240
ton, W. V., Carswell, A., Catling, D. C., Clark, B. C., Duck, T., Dejong,
241
E., Fisher, D., Goetz, W., Gunnlaugsson, H. P., Hecht, M. H., Hipkin, V.,
242
Hoffman, J., Hviid, S. F., Keller, H. U., Kounaves, S. P., Lange, C. F.,
243
Lemmon, M. T., Madsen, M. B., Markiewicz, W. J., Marshall, J., McKay,
244
C. P., Mellon, M. T., Ming, D. W., Morris, R. V., Pike, W. T., Renno,
245
N., Staufer, U., Stoker, C., Taylor, P., Whiteway, J. A., Zent, A. P., 2009.
246
H2 O at the Phoenix Landing Site. Science 325, 58–61.
247
Stillman, D. E., Grimm, R. E., 2011. Dielectric signatures of adsorbed and
11
248
salty liquid water at the Phoenix landing site, Mars. Journal of Geophysical
249
Research Planets 116 (E09005).
250
251
Zent, A. P., 2014. The Phoenix TECP relative humidity sensor. Eigth International Conference on Mars Abstract #1474.
252
Zent, A. P., Hecht, M. H., Cobos, D. R., Campbell, G. S., Campbell, C. S.,
253
Cardell, G., Foote, M. C., Wood, S. E., Mehta, M., 2009. Thermal and
254
electrical conductivity probe (TECP) for Phoenix. Journal of Geophysical
255
Research Planets 114 (E00A27).
256
Zent, A. P., Hecht, M. H., Cobos, D. R., Wood, S. E., Hudson, T. L.,
257
Milkovich, S. M., Deflores, L. P., Mellon, M. T., 2010. Initial results from
258
the thermal and electrical conductivity probe (TECP) on Phoenix. Journal
259
of Geophysical Research 115, E00E14.
260
Zent, A. P., Hecht, M. H., Hudson, T. L., Wood, S. E., Chevrier, V. F., 2012.
261
A Revised Calibration Function for the TECP Humidity Sensor of the
262
Phoenix Mission. 43rd Lunar and Planetary Science Conference Abstract
263
#2846.
264
Zent, A. P., Howard, J., Quinn, R. C., 2001. H2 O Adsorption on Smectites:
265
Application to the Diurnal Variation of H2O in the Martian Atmosphere.
266
Journal of Geophysical Research 106 (7), 14667–14674.
267
7. Figures and Figure Captions
12
Figure 1: Humidity data binned and averaged over 1 hour intervals as a function of local
hour for a) water vapor pressure and b) relative humidity. Colors represent sol spans:
yellow = 1 - 36, blue = 37 - 73, red = 74 - 110, and black = 111 - 152.
13
Figure 2: Pressure and inverse temperature data binned and averaged over 1 hour intervals.
Colors represent LST time spans where red = 000 - 0300, orange = 0400 - 0700, yellow =
0800 - 1100, green = 1200 - 1500, blue = 1600 - 1900, and violet = 2000 - 2300.
14
Figure 3: Phoenix inferred water vapor pressure data compared to simulated surface
temperatures from models (a) and (b) described in text, where the colors represent time
spans and the lines are least squares fits. Blue is for after midnight, red and green are
both for evening hours; however, red is when PH2 O decreases and green is when PH2 O
increases during the evening. Data is restricted for sols when observations were made for
more than 3 consecutive hours during nighttime.
15