Below is a post to sci.astro discusssing evidence there could be ice in the low latitude Tharsis region on Mars.
The evidence of anomalously high water vapor over Tharsis is discussed in this report:

Title: Anomalous H2O Concentrations over Tharsis: Analysis of Equilibrium Models.
Authors: Zent, A. P.; Haberle, R. M.; Murphy, J. R.
Journal: Abstracts of the Lunar and Planetary Science Conference, volume 26, page 1555, (1995)
Full text.

This report reviews suggested explanations that it stems from water in the regolith in Tharsis but doubts that explanation since such water should disappear on relativiely short time scales.
However, if the water is in the form of an *active* glacier then the replenishment of the glacial ice may permit this water to persist long term.
Note also that recent observations suggest there may be currently active glaciers in the low latitude region of Valles Marineris:

ROCK GLACIER-LIKE LANDFORMS IN VALLES MARINERIS, MARS.
Lunar and Planetary Science XXXI.
Full text.

VERY RECENT, WATER-RELATED LANDFORMS ON MARS.
Lunar and Planetary Science XXXII (2001)
Full text.

Such ice in Valles Marineris may be the source for low-lying fogs seen in Noctis Labyrinthus, part of the canyon system in Marineris:

------------------------------------------------------------
NASA SP-441: VIKING ORBITER VIEWS OF MARS
------------------------------------------------------------
- THE ATMOSPHERE -
"Early Morning Clouds in the Tharsis Montes and Valles Marineris Region. Ascraeus Mons and Pavonis Mons are prominently displayed in this mosaic, and dense cloud blankets cling to their northern slopes. High cirrus clouds lie to the west of Tharsis, and waves are visible in the clouds surrounding the peaks. Bands of clouds appearing to have a cellular structure extend north from the canyon, and the areas within and immediately surrounding the chasm exhibit water-ice fogs. [211-5049; 5°S, 105°W]"
"Early Morning Surface Fog. The presence of morning fogs in some crater and channel bottoms is a Viking discovery with possible implications for the future biological exploration of Mars. These early morning views of the Memnonia region were taken one-half hour apart using a violet filter to enhance the contrast of the condensates. The areas marked by arrows are noticeably brighter in the later picture. The fogs indicate specific spots where water is exchanged, probably on a daily cycle, between the surface and the atmosphere. The surface and lower air layers in this region become unusually cold at night because of the thermal properties of the surface. When the surface warms in the morning, it seems that a small amount of water vapor-estimated to be about one-millionth of a meter thick if liquefied is driven off; this vapor recondenses in the atmosphere, which warms more slowly, to form a ground fog of ice particles. [P17487; 13°S,147° W]"
"Early Morning Clouds in Noctis Labyrinthus. Condensate clouds are seen here in early morning in the canyons of Labyrinthus Noctis, which lies at the western end of the equatorial Valles Marineris system. This picture, which covers about 90 000 km2, was made by combining three frames of the same field taken through violet, green, and red filters. Although these clouds lie mainly down inside the canyons, they evidently extend above the walls and spill over some of the surrounding plateau. Like most condensate clouds in the Martian troposphere, they are believed to be composed of water-ice crystals. [P18114, 9°S, 95°W]"
Text and images here.

This Viking era reference suggests these low-lying fogs consist of water ice crystals. However, recent studies of water in clouds suggest supercooling can take place in fogs and clouds to -40C and below. Then such low altitude, low latitude fogs and clouds may consist in part of supercooled liquid water.

Bob Clark

______________________________
______________________________
From: Robert Clark (rgregoryclark@yahoo.com)
Subject: Water/Ice near the equator in Tharsis on Mars.
Newsgroups: sci.astro, alt.sci.planetary, sci.geo.geology, sci.geo.meteorology
Date: 2002-07-03 18:39:44 PST

Andrew Yee <ayee@nova.astro.utoronto.ca> wrote
> ...
> Water-ice clouds continue to become more frequent, especially near
> or over all the major volcanoes, forming a thin cover at southern
> mid-latitudes (see 'C'). Under some conditions, dust storm activity
> warms the atmosphere and inhibits water condensation. However, as
> noted last week, present atmospheric temperatures and water
> distributions (probably in the form of diurnally-precipitated
> frosts that sublime each morning), combined with the slightly dusty
> atmosphere, appear to favor nucleation of mid-atmospheric (at or
> slightly above the tropopause) ice clouds.
>
> ...

This page ascribes the origin of such clouds over the Tharsis volcanoes to rising water vapor on the slopes:

Regional View of the Tharsis Volcanoes
MGS MOC Release No. MOC2-134, 10 June 1999
"The white or bluish-white features are clouds. Clouds are common over the larger Tharsis volcanoes in mid- afternoon. The four largest volcanoes are more than 15 km (9 mi) high. Viewed from Earth by telescope before any spacecraft had visited the planet, astronomers often described a "W-shaped" white cloud over the Tharsis region.
This "W" was actually the result of seeing the combined effects of bright clouds hanging over the Ascraeus, Pavonis, Arsia, and Olympus volcanoes. The clouds result when warm air containg water vapor rises up the slopes of each volcano, cools at the higher altitude, and causes the water vapor to freeze and form a cloud of ice crystals."
http://www.msss.com/mars_images/moc/6_10_99_tharsis/

However, there has been observed anomalously high amounts of water
vapor in the Tharsis region. What is the origin of this water vapor?

Title: Remote sensing evidence for regolith water vapor sources on Mars.
Authors: Huguenin, R. L.; Clifford, S. M.
"A survey of ground-based and spacecraft observations at visible wavelengths of condensate activity in the Martian atmosphere is presented. The observations reveal that Solis Lacus has been a center of condensate activity
throughout the observational record, along with sites in Noachis-Hellespontus and in border regions in Syrtis Major. The condensate activity did not appear to be
controlled by the distribution of aerosol dust, surface albedo, thermal inertia, or topography. Rather, the patterns of cloud and fog occurrence, together with
observed and inferred surface winds, point to these sites as primary H2O vapor source regions for condensate
development. Specifically, it is proposed (1) that the unusual distribution of condensates around Solis Lacus at
the onset of the 1973 dust storm may have resulted from the transport of moisture from the central plume site in
northern Solis Lacus (near Sinai Planum) by prevailing daytime winds and nighttime drainage winds, (2) that the
recurrent winter Tharsis and Hellas condensate features may have developed from MOISTURE DERIVED PRIMARILY FROM THE UPWIND SOLIS LACUS AND NOACHIS-HELLESPONTUS SOURCE REGIONS, respectively, and (3) that the recurrent summer Elysium clouds may have developed as a result of MOISTURE FROM UPWIND SYRTIS MAJOR BORDER SITES. The estimated
outgassing rate from Solis Lacus during 1973 was ~10^-1 km^3 yr^-1, and a similar rate is suggested for the Noachis-Hellespontus site. These rates would probably not be high enough to be detected by the Viking MAWD experiment, but they extrapolate to the possile loss of a meter of ice from the sites in only 10^3 years. IT IS PROPOSED THAT THE ICE IS PRESENT IN THE SURFACE LAYER AND THAT IT COULD UNDERGO MELTING DURING A PORTION OF THE DAY OVER AN APPRECIABLE FRACTION OF THE YEAR. RADAR, THERMAL, AND WATER VAPOR DATA WOULD BE CONSISTENT WITH THE
POSSIBLE OCCURRENCE OF SOIL MOISTURE AT THESE SITES."
[emphasis added]
Journal of Geophysical Research, vol. 87, no. B12, Nov. 30, 1982, p. 10227-10251.

This paper proposes the higher amount of water vapor in Tharsis stems from the Solis Lacus and Noachis Hellespontus regions. However, since Tharsis is a volcanic region and recent evidence suggests Mars should still be volcanically active it may be that the water vapor
results from volcanic outgassing.
However, another possibility is suggested by the fact that the amount of water vapor varies diurnally. Perhaps this water vapor arises from the regolith in Tharsis. Johann Helgason has argued that there are currently active glaciers in Tharsis:

Does Mars Hide Vast Water Deposits?
by Johann Helgason
Reykjavik - June 10, 2000
http://spacedaily.com/news/mars-water-99b.html

The recent GRS reports that show ice deposits in the Tharsis low-latitude region as well as in the expected mid to high latitudes on Mars is supportive of this idea.

Bob Clark
_______________________________
_______________________________
Discussion thread.