ment 
| and 
ring 
indle 
f the 
nage 
inate 
on is 
Cable 
> our 
| and 
sults 
ilues 
lowy 
in is 
on of 
ding 
The 
gery 
thod 
| and 
r the 
) the 
e on 
ithm 
n to 
t, the 
aken 
iakes 
ht of 
M) is 
th of 
d on 
:dge- 
btain 
et al. 
mage 
s not 
yased 
ected 
ed at 
'eady 
also 
able. 
jased 
sition 
anted 
lency 
  
     
P 
Figure 1. (A) Subset of NAC image M135433752LE covering a 
boulder field situated around -30.4°lat and 322.5°lon 
(B) Groups of pixels detected as bright faces of boulders on the 
image (colors correspond to ranges of size) 
2.5 Risk Maps 
The computation of risk maps for one specified landing area is 
mainly based on three hazards: slope, roughness and shadow. 
Slope and roughness hazard maps are computed from the high 
resolution DTM while the shadow map is produced by the 
illumination module (s. next section). The risk map 
computation consists of the combination of all hazards in one 
binary (safe/unsafe) landing risk map at a certain alpha-level. 
The alpha level is the probability that the threshold is 
outmatched (toward infinite) using a normal probability 
distribution function centered on the current value and with a 
dispersion defined by the error map (variance). Each hazard 
map is transformed into a binary risk-score map at a certain 
alpha-level according to user-defined thresholds. The error on 
the hazard is taken into account in the comparison between the 
current value of the hazard and the threshold. The final landing 
risk map is computed by multiplication of individual risk score 
maps and taking into account the landing position uncertainty 
(Bonfiglio et al. 2011). 
2.6 Illumination Maps 
Before being able to generate any illumination product for a 
target area, the illumination module generates a horizon mask 
using a ray-tracing algorithm. On the basis of a DTM it 
computes in each pixel of the target area and for each direction 
the biggest elevation angle of an obstacle along this direction. 
Using the horizon mask corresponding to a specified time (i.e. a 
specified sun direction), an instantaneous illumination map 
showing the proportion of the solar disk visible in each pixel 
and an incoming flux map providing the irradiance in Watt/m? 
at a specified time can be computed. 
Maps characterizing the illumination conditions on a certain 
period can also be produced like the accumulated Sun 
illumination fraction map (providing for each pixel the Sun 
illumination fraction for a period of analysis), the longest quasi 
continuous illumination period (LQCIP) or the longest period of 
darkness (LPOD), and others (Vanoutryve et al. 2010). 
2.7 Temperature Maps 
This module can be subdivided into two parts. The first one 
allows the generation and update of one pre-processed set of 
3D-grid data containing an aggregation of all brightness 
temperature measurements from LRO DLRE (Paige et al. 2010) 
in three thermal channels (7, 8 and 9) of the instrument since 
the beginning of the observations. These three brightness 
temperature values provided in DLRE RDR tables in the NASA 
PDS supply the 3D-grid established with a predefined time step 
and spatial resolution for a typical lunar day on the whole moon 
surface, excluding the effect of eclipses and other incoherent 
data (filtered using quality flags). A temporal filter is applied to 
smooth erratic values and linear interpolation is computed for 
the spatiotemporal pixels where no DLRE brightness 
temperature data is directly available. Surface temperature is 
considered equivalent to brightness temperature because the 
moon soil emissivity in the wavelengths of the three channels is 
considered equal to 1. This condition is verified by the high 
correlation between the different channels (Paige et al. 2010). 
The second part of the temperature module allows the 
generation of two kinds of surface temperature products from 
this dataset. A low resolution temperature map can be obtained 
for one specified time (interpolated from the dataset described 
above) and the mean diurnal variation of the surface 
temperature at one specified point can be plotted in a diagram. 
2.8 Reflectance Maps 
Reflectance maps are computed from LROC NAC or WAC 
(Wide-Angle Camera) radiance images. The reflectance can be 
understood as the ratio (value between 0 and 1) between the 
power re-emitted by the surface (in all directions of the 
hemisphere above the observed target) and the received flux 
from the Sun on this surface (7 the irradiance). It does not 
depend on the topography, nor on the illumination conditions; it 
only depends on the physical characteristics of the surface. 
Sun position, sensor position and surface topography must be 
known to compute reflectance from a set of radiance images. 
Several reflectance models can be used to estimate the 
reflectance from radiance images (Lambertian, Lunar-Lambert 
and Hapke models are forescen for LandSAfe). 
475