Analysis of Soil Variability Measured with a Soil
B. HANQUET, D. SIRJACOBS AND M.-F. DESTAIN email@example.com
Gembloux Agricultural University, Unite
canique et Construction, Passage des De
B-5030 Gembloux, Belgium
Centre de Recherches Agronomiques, De
partement Production Ve
tale, Rue du Bordia 4,
B-5030 Gembloux, Belgium
Libre de Bruxelles, Service de Me
canique des sols, Avenue F. Roosevelt 50,
B-1050 Bruxelles, Belgium
Abstract. In the context of precision agriculture, the knowledge of soil strength variability at the ﬁeld
scale may be useful for improving site-speciﬁc tillage. Moreover, rapid and accurate sensing methods for
soil physical properties determination would favourably replace labour-intensive, time-consuming and
expensive soil sampling and analysis. This study aims at validating conclusions of a previous study which
was conducted to develop and test a soil strength sensor in ﬁeld conditions. The coupled acquisition of the
sensor’s signals and the corresponding DGPS positions allowed establishment of maps for the three
measured outputs, namely the horizontal force (F
), the vertical force (F
) and the moment (M
). In order
to study the relationships between measured forces and soil physical parameters, a series of soil properties
were measured on soil cores collected in 10 reference plots. Signiﬁcant correlations were found between F
and the average resistance to cone penetration at 25 cm depth (r ¼ 0.95) and between F
and average soil
moisture at 30 cm depth (r ¼ )0.95). These relationships were similar to those found in the ﬁrst study.
This sensing method proved its capability to characterise within-ﬁeld soil variability.
Keywords: sensor, mapping, soil variability, soil strength, soil physical properties
Among crop production factors, the soil is obviously one of the most important.
Therefore, within the context of precision farming, the knowledge of its physical and
mechanical properties, as well as the spatial variability of these properties, is essential
as decision support information for modifying cultural operations. Up to now, data
collection on soil is mostly made by grid-sampling to determine, e.g., texture, pF,
water content, water inﬁltration rate, bulk density, cone index, organic matter and
shear strength. Measurements of these properties are labour-intensive, time-con-
suming and expensive. Thus, the development of sensors suited to quantify soil
properties at the scale required for accurately mapping within-ﬁeld variations is a
necessity in order that precision agriculture can be widely practised (Staﬀord, 2000).
Precision Agriculture, 5, 227–246, 2004
Ó 2004 Kluwer Academic Publishers. Manufactured in The Netherlands.