SATELLITE OBSERVATION OF BARE SOILS FOR THEIR

AVERAGE DIURNAL ALBEDO APPROXIMATION

Jerzy Cierniewski

Department of Soil Science and Remote Sensing of Soils, Institute of Physical Geography and Environmental Planning,

Adam Mickiewicz University, 61-680 Poznań, Poland

Keywords: Soil, Roughness, Albedo, Satellite Orbit.

Abstract: This study explores the diurnal variation in broadband blue-sky albedo (

) of soils with respect to their

roughness. Uncultivated soils and cultivated ones, after ploughing, harrowing and rolling were studied in

Israel and Poland. The relation between

of the surfaces and the solar zenith angle allowed to predict the

diurnal

variation of the surfaces located at a given latitude at any date, and calculate the optimal time T

related to their average diurnal albedo

observation. This procedure was used to assess the usefulness of

satellites on the sun-synchronous orbits for approximation of

for the moderately rough bare soil

surfaces, located between the latitude angles of 75° S to 75° N, within an error lower than ±2%. It was

found that the satellites on the orbits crossing the Equator at 10:30, such as the MODIS, and like the SPOT

and IRS IC, are not very useful for that, while the NOAA-15 on the orbit crossing the Equator at 7:30 is

much more useful. The best dates for the collection of data with this satellite are 16 April and 28 August,

while the worst date is June 22.

1 INTRODUCTION

The Earth’s surface shows variation in its reflected

radiation due to the direction of irradiating solar

energy and the direction along which the reflected

energy is viewed by ground, air-born and satellite

sensors. The reason for this variation is the

irregularities of the surface, which produce shadow

areas, where the solar beams do not directly reach all

surface areas. The radiation leaving shaded areas is

many orders-of-magnitude smaller than radiation

reflected from sunlit fragments. Reflectance of the

Earth’s surface is usually highest from the direction

which gives the lowest proportion of shaded

fragments. For example, bare soils with irregularities

caused by the soil texture, aggregates and micro-

relief configuration usually display a backscattering

reflectance peak towards the Sun, and decreasing

reflectance in the direction away from the peak

(Milton and Webb, 1987; Cierniewski et al., 2004).

Desert surfaces can have both a backscattering and

forward-scattering character (Shohsany, 1993). They

display maximum reflectance in the extreme

forwardscatter direction near horizon if they are

relatively smooth with a strong specular behaviour

(Coulson, 1966).

The bidirectional reflectance is defined as the

fraction of incident radiation that is reflected from a

surface along a given direction. Its value

corresponds to the precisely specific direction of the

surface illumination. Another physical quantity

describing a reflectance of a surface is the albedo. It

integrates the surface reflectance over all view

angles and is defined as the fraction of the incident

solar short-wave radiation (0.3-3μm) that is reflected

from a surface. The upwelling and downwelling

radiations are integrated over the whole hemisphere

(Schaepman-Strub et al., 2006). Martonchik et al.

(2000) recommend using the terms broadband

albedo or the narrowband (spectral) albedo if the

albedo characterizes the entire solar short-wave

spectrum or only a part of it, respectively. The blue-

sky albedo describes the albedo measured under

field conditions, where a surface is illuminated by

both direct solar irradiance and diffuse irradiance,

scattered by the atmosphere (Baret et al., 2005). In

many parts of the Earth, the reflectance of the land

surface changes seasonally. However, annual

variation of albedo is usually smaller than its daily

variation. The albedo, characterizing the intrinsic

properties of a surface, such as the surface

bidirectional reflectance, varies with solar zenith

229

Cierniewski J..

SATELLITE OBSERVATION OF BARE SOILS FOR THEIR AVERAGE DIURNAL ALBEDO APPROXIMATION.

DOI: 10.5220/0003903902290234

In Proceedings of the 1st International Conference on Sensor Networks (SENSORNETS-2012), pages 229-234

ISBN: 978-989-8565-01-3

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

angle (Liang et al. 1999). Kondratyev (1969)

mentioned that, during the morning when the sun

elevation increased from 10°

to 60°, the albedo of

dry rocky and loamy soil surfaces decreased from

0.22 to 0.14 and from 0.34 to 0.21, respectively.

Pinty et al. (1989), Eckardt (1991), Lewis and

Barnsley (1994), Matthias et al. (2000) and Wang et

al. (2005) confirmed that the albedo of cultivated

and uncultivated desert surfaces varied significantly

at relatively high solar zenith angles.

Satellite radiometers are not able to measure

directly the broadband blue sky Earth’s surface

albedo. The anisotropic, non-Lambertian distribution

of the reradiation of soil and other Earth’s surfaces

must be inferred through a series of manipulations of

the raw remote sensing reflectance data (Maurer,

2002). Only the data representing cloud-free pixels

are analyzed. Since the satellite instruments measure

the radiance at the top of the atmosphere, and we

intend to estimate the albedo of the Earth’s surface,

a correction is required in order to account for the

effect of the intervening atmosphere. Because most

satellites collect the Earth’s surface data only at one

or a few directions inside their small field-of-view,

these data are recomputed with consideration of the

bidirectional reflectance of Earth objects. Lastly,

since satellites measure the Earth’s radiation in a

number of separate narrowband channels, an

extrapolation of the narrowband albedo values to

their broadband values is necessary.

The broadband albedo is used for modelling

environmental biophysical processes associated with

the energy transfer between soil, vegetation and

atmosphere, as well as for studying climate changes

at regional and global scales. Because the processes

are analyzed in the diurnal cycle, as well as over

wider time ranges, such as monthly, seasonal and

year-long, the average diurnal surface albedo seems

to be a helpful basis for the modelling. Sellers et al.

(1995) determined the accuracy requirement of

albedo for the global models as ±2%. It seems that

more attention should be focused on the time of the

satellite data acquisition because: i) the albedo

clearly vary during the day, ii) its value is required

with such a small error, and iii) the algorithms used

to convert the reflectance of a surface to its albedo,

relying on a semi-empirical approach (Olsen et al.,

2003, Zhou et al., 2003,Tsvetsinskaya et al., 2006),

are aimed at an effective elimination of the influence

of the non-equal distributed radiation from the

surface during a satellite passing.

This paper shows how the albedo of bare soil

varies as the solar zenith angle function depending

on their roughness. Examples of the broadband blue-

sky soil albedo sets were used, measured from the

ground level from midday to sunset. These relations

allowed to predict the albedo variation of bare soil

surfaces located at a chosen latitude during any day

of the year. The paper shows in which local solar

time the albedo of the moderately rough soil

surfaces located between the latitude angles of 75°

the Northern and Southern Hemispheres represents

its diurnal averaged value within an error lower than

±2%. We consider how it is possible to observe a

soil surface from satellites on sun-synchronous

orbits in the optimal time close to the moment when

its albedo reaches the average diurnal value with that

low error. The usefulness of a satellite placed on the

orbit close to that being optimal, as well as satellites

which most often have been used to the albedo

approximation, is analyzed here.

2 METHODS

The paper reports the relation between the

broadband blue-sky albedo

of soil surfaces in

Israeli Negev desert near Sede Boker (30° 51’26”N,

34°47’09”E) and the solar zenith angle

. There, it

was possible to collect the data relating to the soils

developed from a loessial substrate with an

extremely high diversity of their roughness,

including extremely smooth uncultivated soil

surfaces, as well as moderate and very rough

cultivated surfaces after shallow and deep

ploughing, respectively (Fig. 1). These studies have

been continued in Poland on soil surfaces in the

western part of Poland near Poznań (52°34’57”N,

16°38’49.23”E; 52°28’48”N, 16°49’49.36”E). These

include cultivated soils (Luvisols according the

World Reference Base for Soil Resources),

developed from sands and sandy loams. Figure 1

shows three examples of such soils that were shaped

by ploughing, harrowing and rolling, respectively.

The α values of all surfaces were measured by an

albedometers LP PYRA 06 in a spectral range of

0.335–2,200 μm using data loggers. Shape of the

surfaces was controlled by a 3D laser scanner

Konica-Minolta VIVID-910. The texture and the

organic carbon content of the soil surface material

was controlled in the laboratory using a hydrometer

and Walkley Black’s method, respectively (Sparks et

al., 1996).

3 RESULTS AND DISCUSSION

This paper focuses on the most important results

SENSORNETS 2012 - International Conference on Sensor Networks

230

Figure 1: Soil surfaces tested in Israel (A-C) and Poland

(D-F).

obtained during tests carried out in Israel and

Poland.

The measurements of the albedo were carried out

in Israel on August 2008, and those in Poland from

May to September, 2011, under clear sky conditions

from solar local noon to sunset.

The relation between

and

, derived from the

studies in Israel (Cierniewski et al., 2012) and

Poland clearly shows that

of a bare soil strongly

depends on its surface roughness (Fig. 2).

The

curves for the smooth surfaces (the uncultivated

surface and the cultivated ones after rolling) are

consistently higher than those of the very rough

surfaces (deeply ploughed). The curves related to the

extremely rough surfaces do not rise at

angles

lower than 80°, while for the extremely smooth ones

they increase throughout the analyzed

range,

rising strongly at

higher than 75°. The shape of

the curves clearly affects the

value at which the

reaches its average value.

The above relations, expressed by curvilinear

functions, were used to compute instantaneous

values of

for surfaces of a given roughness,

located at a chosen latitude during any day of the

year. They were also used to find the optimal time

relating to their average diurnal albedo

observation. The T

value was defined as the local

solar time when the

value, acquired by an

instantaneous observations best represents the

value for an given day (from almost sunrise to

sunset, i. e., when the

does not exceed 85°).

Figure 2: Variation of the broadband blue-sky albedo (

)

of the studied surfaces as a function of the solar zenith

angle

. Vertical red lines mark

values that relate to the

average diurnal albedo

of the surfaces.

The semi-diurnal distributions of

and

of the

moderately rough soil surface (such as that tested in

Israel) are presented in Figure 3. The panels

represent selected days in the Northern hemisphere

(NH) to: the astronomical spring (21 March) and

autumn (23 September) equinoxes and the

SATELLITE OBSERVATION OF BARE SOILS FOR THEIR AVERAGE DIURNAL ALBEDO APPROXIMATION

231

Figure 3: Morning semi-diurnal distributions of the

broadband blue-sky albedo

for the moderately rough

soil surfaces in the latitude (L) and local solar time (LST)

functions for the chosen dates. The bold isoline on the left-

hand side of the graphs shows extreme values of α for the

solar zenith angle

= 85°, and bold dotted line - the

average diurnal

value for a given L. NH and SH

represent the Northern and Southern Hemispheres,

respectively.

beginning of the astronomical summer (22 June) and

winter (22 December), and additionally other days

that have a different pattern than mentioned above

(23 February, 16 April, 28 August and 19 October).

Each panel show contour lines (emphasized with

shading) of the

values as a function of the local

solar time and the latitude. The bold dotted line

shows the

values for a given latitude

(Cierniewski and Gdala, 2010).

Figure 4 gives the opportunity to observe the

moderately rough soil surfaces (as representing the

most common cultivated bare areas) at the time

close to the moment when their albedo viewed from

satellites on sun-synchronous orbits reaches the

average diurnal value with error lower than ±2%.

The following dates are chosen: 21 March, 16

April, 7 May and 22 June, using the examples of the

satellite NOAA-15 and MODIS, crossing the

Equator at 7:30 and 10:30 morning, respectively.

The orbit of the NOAA-15, as one of the few

civilian remote sensing satellites with the

descending node at so early local solar time, allows

to obtain these rough soil albedo data with the

acceptable error in limited ranges of the latitude. The

ranges are described by intersection points of the

two kinds of lines, describing limits of the

acceptable error ±2% of the data and the time of the

satellite passage. These latitude ranges are: 30°N-

50°S for 21 March, 75°N-15°S for 16 April, 60°N-

5°S for 7 May and 25°N-5°S for 22 June. The widest

range, larger than 90°, was found for the 16 April

(and correspondingly for 28 August) and the

narrowest one, 30°, for June 22.

The data obtained from the MODIS (like the

SPOT and the IRS IC), which have been often used

for the albedo approximation, can be much less

useful for this purpose. As Figure 4 shows, one can

obtain sufficiently precise results only in the very

narrow ranges of the latitude, not wider than 5°, and

moreover in the very high latitudes positions, higher

than 65°S for 16 April, 60°S for 7 May and 50°S for

22 June.

4 CONCLUDING REMARKS

The broadband blue-sky albedo

variation of a bare

soil as a function of the solar zenith angle

clearly

depends on the soil surface roughness. The curves

describing this relation for the extremely rough

surfaces almost do not rise at

angles lower than

80°, while the curves for the extremely smooth

surfaces increase throughout the analyzed

ranges,

rising strongly at

higher than 75°. The shape of

the curves clearly affects the

value at which

SENSORNETS 2012 - International Conference on Sensor Networks

232

Figure 4: Time in the hour scale of local solar time (LST),

when the average diurnal albedo of the moderately rough

soil

(green line), predicted for the chosen dates varying

with the latitude (L), is available to be approximated with

the error lower than ±2% (dotted lines). The red and

orange lines depict the time the NOAA-15 and MODIS

passage. NH and SH represent the Northern and Southern

Hemispheres, respectively.

instantaneous

reaches its average value. The

relations between

and

, expressed by curvilinear

functions, allow to predict the

variation of bare

soils located at a given latitude at any date, and

calculate the optimal time T

relating to their

average diurnal albedo

observation.

Such a procedure allowed to asses the usefulness

of satellites on the sun-synchronous orbits for

approximation the

of the most common

cultivated bare soil areas within an error lower than

±2%. The moderately rough soil surfaces, located

between the latitude angles of 75°

in the Northern

and Southern Hemispheres were chosen for this

assessment. It was found that the satellites on the

orbits crossing the Equator at 10:30, such as the

MODIS, and like the SPOT and IRS IC, are not very

useful for that, while the NOAA-15 on the orbit

crossing the Equator at 7:30 is much more useful.

Tho first group of the satellites is only useful for the

surfaces located in very high latitudes and not for all

dates. The latter satellite is useful within wide

latitude ranges in lower latitudes. The best dates for

the collection of data with this satellite are 16 April

and 28 August, while the worst date is June 22.

Approximation of the Earth’s surface albedo via

satellite data, collected closely to the time T

when

the soil surfaces reaches

value, will probably

result in a significantly reduced error in the

calculation of the albedo. The semi-empirical

approach, used to convert the soil reflectance to its

albedo, would be simpler. It would be only limited

to correction of the reflectance distribution relative

to a specific direction of a satellite viewing.

ACKNOWLEDGEMENTS

This work was supported by the Polish Ministry of

Science and Higher Education carried out under the

Project NN306600040. The author wishes to thank

Prof. Bajorski for the language revision of the

manuscript, as well as Dr. Cezary Kaźmierowski,

Dr. Sławomir Królewicz, Dr. Jan Piekarczyk,

Krzysztof Kuśnierek, MSc, and Michał Rymaniak,

MSc for their help in the measurements and

preparation of the figures.

REFERENCES

Baret, F., Schaaf, C., Morisette, J. and Privette, J. (2005).

Report on the Second International Workshop on

Albedo Product Validation. The Earth Observer, 17,

13–17.

Cierniewski, J., Gdala, T. and Karnieli, A. (2004). A

hemispherical-directional reflectance model as a tool

for understanding image distinctions between

SATELLITE OBSERVATION OF BARE SOILS FOR THEIR AVERAGE DIURNAL ALBEDO APPROXIMATION

233

cultivated and uncultivated bare surfaces. Remote

Sensing of Environment, 90, 505–523.

Cierniewski, J. and Gdala T. (2010). Calculating the

optimal time when albedo approximates its daily

average: an example using soil surfaces with various

roughnesses at different latitudes. International Jour-

nal of Remote Sensing, 31, 2697-2708.

Cierniewski, J., Karnieli, A., Kuśnierek, K. and Herrmann,

I. (2012) Approximation the average daily surface

albedo with respect to soil roughness and latitude.

International Journal of Remote Sensing (accepted for

publication).

Coulson, K. L. (1966). Effect of reflection properties of

natural surfaces in aerial reconnaissance. Applied

Optics, 5, 905-917.

Eckardt, M. (1991). Albedo changes and satellite

observations of the Earth. In Climate and Global

change – Proceedings of the European School of

Climatology and Natural Hazards Course, held in

Arles/Rhone, France, from 4 to 12 April 1990.

Brussels, Lxemburg: ECSC-EEC-EAEC, 123-137.

Kondratyev, K. Y. (1969). Radiacjonnyje Charakteristiki

Atmosfery i Zemnoy Poverkhnosti. Leningrad: Gidro-

meteorologicheskoye Izdatelstvo.

Lewis, P. and Barnsley, M. J. (1994). Influence of the sky

radiance distribution on various formulations of the

Earth surface albedo. In Proceedings Mesures

Physiques et Signatures en Teledetection, Val d’Isere,

France, 17–21 January, 1994, 707–716.

Liang G, S., Strahler, A. H. and Walthall, C. (1999).

Retrieval of land surface albedo from satellite

observations: a simulation study. Journal of Applied

Meteorology, 38, 712–725.

Martonchik, J. V., Brugge, C. J. and Strahler, A. (2000). A

review of reflectance nomenclature used in remote

sensing. Remote Sensing of Environment, 19, 9–20.

Matthias, A. D., Fimbres, A., Sano, E. E., Post, D. F.,

Aciolly, L., Batchily, A, K., and Ferreira, L. G.

(2000). Surface roughness effects on soil albedo, Soil

Science Society of America Journal, 64: 1035-1041.

Maurer J. (2002). Retrieval of surface albedo from space.

Retrived from: http://www2.hawaii.edu/~jmaurer/.

Milton, E. J. and Webb, J. P. (1987). Ground radiometry

and airborne multispectral survey of bare soils.

International Journal of Remote Sensing, 18, 3-14.

Olsen K. W., Bonan, G. B., Schaaf, C., Gao, F., Jin, Y.

and Strahler, A. (2003). Assessment of global climate

model land surface albedo using MODIS data.

Geophysical Research Letters, 30, 1443, doi: 10.1029/

2002GL016749.

Pinty, B., Verstratte, M. M. and Dickinson, R. E. (1989).

A physical model for predicting bidirectional

reflectance over bare soil. Remote Sensing of

Environment, 27, 273–288.

Schaepman-Strub, G., Schaepman, M. E., Painter, T. H.,

Dangel, S. and Martonchik, J. V. (2006). Reflectance

quantities in optical remote sensing – definitions and

case studies. Remote Sensing of Environment, 45, 15–

27.

Shoshany, M. (1993). Roughness-reflectance relationship

of bare desert terrain; An empirical study. Remote

Sensing of Environment, 103, 27–42.

Sellers, P. J., Meeson, B. W., Hall, F. G., Asrar, G.,

Murphy, R. E., Schiffer, R. A., Bretherton, F. P.,

Dickinson, R. E., Ellingson, R. G., Field, C. B.,

Hummerich, K. F., Justice, C. O., Melack, J. M.,

Roulet, N. T., Schimel, D. S. and Try, P. D. (1995).

Remote Sensing of the land-surface for studies of

global change: models-algorithms-experiments.

Remote Sensing of Environment, 51, 3-26.

Sparks, D. L., Page, A. L., Helmke, P. A., Loeppert, R. H.,

Soltanpour, P. N., Tabatabai, M. A., Johnston, C. T.

and Summer, M. E. (1996). Methods of Soil Analysis.

Part 3. Madison: Soil Science Society of America and

American Society of Agronomy.

Tsvetsinskaya, E. A., Schaaf, C.B., Gao, F., Strahler, A.

H. and Dickinson, R. E. (2006). Spatial and temporal

variability in moderate resolution imaging

spectroradiometer derived surface albedo over global

arid regions. Journal of Geophysical Research, 111,

D20106, doi:10.1029/2005JD006772.

Wang, Z., Barlage, M. and Zeng, X. (2005). The solar

zenith angle dependence of desert albedo. Geophysical

Research Letters, 32, doi: 10.1029/2004GL021835.

Zhou, L., Dickinson, R. E., Tian, Y., Zeng, X., Dai, Y.,

Yang, Z.-L., Schaaf, C. B., Gao, F., Jin, Y., Strahler,

A., Myeneni, R. B., Yin, H., Wu, W. and Shaikh, M.

(2003). Comparison of seasonal and spatial variations

of albedos from Moderate-Resolution Imaging

Spectroradiometer (MODIS) and Common Land

Model. Journal of Geophysical Research, 108, 15-1-

15-20.

SENSORNETS 2012 - International Conference on Sensor Networks

234