If only data density were playing a role, topographic information would have had
the highest weighting in terms of developing the map, which was the case, but
radiometric data had a slightly higher influence. This was due to the following
reasons: (1) radiometric data allowed the direct identification of RSGs, while
(2) topographic data were a surrogate for the climatic water balance (ratio of rainfall
and evapotranspiration as an indicator for the leaching rate), allowing us to establish dominance rules depending on the parent material and geographic position. In
the Bor Krai limestone area; for example, Acrisols dominate over Alisols with
increasing topographic height, while Cambisols dominate over Acrisols in the
highest landscape positions in the granitic Mae Sa watershed. In conclusion, data
density plays an important role when using the random forest approach to develop
soil maps, as does the discriminative quality of the data used. This exercise showed
that in this respect, gamma-ray spectrometry has a good potential, which should be
further exploited in the future.
2.2.7 Conclusions and Outlook Regarding Gamma-Ray
Spectrometry
The soil mapping studies here showed that the dominant and characteristic soil
forming process in the northern Thai highlands is clay illuviation. Since gammaray signals integrate over only a limited depth, clay illuviation changes the signal
of the soil profiles due to different clay mineral compositions, by decreasing the
K concentration of the topsoil to different degrees. As a consequence, the K-Th
ratio measured at the surface can be used to distinguish the dominant Reference
Soil Groups (Acrisols, Alisols). At the soil profile scale, the K gamma-ray signal
depth gradient can be reconstructed by combining the clay concentration, bulk
density and potential cation exchange capacity of the clay fraction. At the
landscape scale, >80 % of soil augerings can be attributed to WRB Reference
Soil Groups using field methods and with hand-held gamma-ray spectrometry
included, the latter thus reducing the need for and costs of laborious analytics.
A systematic signal shift occurs between ground-based and airborne
measurements. Therefore, algorithms for soil mapping have to be adapted
when shifting between scales. At the regional scale, gamma-ray information
can be seen as the most important criterion to use for establishing concept soil
maps in the northern Thai highlands. In conclusion, gamma-ray spectrometry can
be used to assist in Reference Soil Group identification at the watershed scale,
and has potential on a regional scale too. Though there are indications that
general algorithms might be used for clay illuviation type RSG mapping worldwide, not all factors that influence the quantitative gamma-ray signal are understood; therefore, more deterministic studies are required.
48
K. Stahr et al.
the highest weighting in terms of developing the map, which was the case, but
radiometric data had a slightly higher influence. This was due to the following
reasons: (1) radiometric data allowed the direct identification of RSGs, while
(2) topographic data were a surrogate for the climatic water balance (ratio of rainfall
and evapotranspiration as an indicator for the leaching rate), allowing us to establish dominance rules depending on the parent material and geographic position. In
the Bor Krai limestone area; for example, Acrisols dominate over Alisols with
increasing topographic height, while Cambisols dominate over Acrisols in the
highest landscape positions in the granitic Mae Sa watershed. In conclusion, data
density plays an important role when using the random forest approach to develop
soil maps, as does the discriminative quality of the data used. This exercise showed
that in this respect, gamma-ray spectrometry has a good potential, which should be
further exploited in the future.
2.2.7 Conclusions and Outlook Regarding Gamma-Ray
Spectrometry
The soil mapping studies here showed that the dominant and characteristic soil
forming process in the northern Thai highlands is clay illuviation. Since gammaray signals integrate over only a limited depth, clay illuviation changes the signal
of the soil profiles due to different clay mineral compositions, by decreasing the
K concentration of the topsoil to different degrees. As a consequence, the K-Th
ratio measured at the surface can be used to distinguish the dominant Reference
Soil Groups (Acrisols, Alisols). At the soil profile scale, the K gamma-ray signal
depth gradient can be reconstructed by combining the clay concentration, bulk
density and potential cation exchange capacity of the clay fraction. At the
landscape scale, >80 % of soil augerings can be attributed to WRB Reference
Soil Groups using field methods and with hand-held gamma-ray spectrometry
included, the latter thus reducing the need for and costs of laborious analytics.
A systematic signal shift occurs between ground-based and airborne
measurements. Therefore, algorithms for soil mapping have to be adapted
when shifting between scales. At the regional scale, gamma-ray information
can be seen as the most important criterion to use for establishing concept soil
maps in the northern Thai highlands. In conclusion, gamma-ray spectrometry can
be used to assist in Reference Soil Group identification at the watershed scale,
and has potential on a regional scale too. Though there are indications that
general algorithms might be used for clay illuviation type RSG mapping worldwide, not all factors that influence the quantitative gamma-ray signal are understood; therefore, more deterministic studies are required.
48
K. Stahr et al.