Visible–NIR reflectance: a new approach on soil evaluation
Introduction
Soil scientists have been challenged and aided during the past few decades by the simultaneous evolution and revolution in methods and instrumentation for soil data acquisition and modeling. A wealth of new soil information has become available to many countries (Baumgardner, 1999). However, some of the information from soil maps are of uneven quality, often outdated and sometimes wrong (Bouwman, 1990). In Brazil, soil surveys are incomplete and mostly developed for great scales (Oliveira, 1988), making difficult land use planning (Fiorio et al., 2000) and precision agriculture systems (Moran et al., 1997). Until today, the situation has not changed significantly (Prado, 1997). Additionally, the relationship between soil survey and soil tillage should be considered for a better soil management Demattê, 1999a, Demattê, 1999b.
Soil surveys are usually performed using field and laboratory conventional techniques (Soil Survey Staff, 1998). However, computational facilities and emerging technologies, such as remote sensing (RS) and geographical information systems (GIS), can be used to assist these surveys (Usery et al., 1995). The basis of RS is that reflected energy interacts with all constituents of soils, allowing quantification and discrimination between them. Ground sensors have been used Stoner and Baumgardner, 1981, Galvão and Vitorello, 1998 and validated the importance of SR for soil characterization (Demattê, 2002). Huete (1996) observed that ground data can be used to validate satellite data and both can be used to assist soil survey. However, problems such as signal-to-noise ratio from satellite data and the relatively small spectral channels have to be considered (Ben-Dor, 2002).
Studies on RS using ground sensor have usually been performed to investigate the relationship between soils and their attributes Formaggio et al., 1996, Madeira Netto, 1996 through descriptive (Stoner and Baumgardner, 1981) and quantitative Ben-Dor and Banin, 1995, Liang and Townshend, 1996 evaluations. The importance of spectral data on soil surveys has been shown; however, detailed information on how to use SR in soil surveys are lacking. Considering the importance of soil surveys for agriculture and environment, it is imperative to improve new methodologies using spectral data (Ben-Dor et al., 1999).
As pointed out by Baumgardner (1999), soil scientists are challenged to integrate available technologies to provide more objective decision-support systems in the management and monitoring of soil resources. Therefore, the objective of this work was to develop a laboratory spectral methodology (450–2500 nm measurements) to assist soil surveys, made primarily for soil tillage. Our hypothesis was that soils along a toposequence would present differences in SR due to their attributes and parent materials. Also, we expected that the simultaneous interpretation of spectral curves from different depths would allow soil classifications.
Section snippets
Characterization of the studied area and soils
The studied area was located in Piracicaba, São Paulo State, Brazil (longitude 47°31′00″ to 47°34′00″; latitude 22°39′00″ to 22°42′00″) and covers 350 ha. Eight soils were identified in the area and classified, according to Soil Survey Staff (1998), as: typic haplorthox (Latossolo Roxo, LR); typic haplorthox (more red, Latossolo Vermellho-Escuro, LE), typic paleudult (Podzólico Vermelho Amarelo, PV), rhodic paleudult (PE), typic paleudult (Podzólico Vermelho Escuro, PE), typic distrochrept
Spectral characterization and discrimination of the soils
Soil designations and its physical and chemical classification are presented in Table 1. Normally, the LR soil is located in a flat or gently rolling landscape, with microaggregation and without textural gradient (Fig. 4), deep weathered and clayey soil (Table 1). LE is very similar to LR and differs in the less total iron content. PE is characterized by an ochric epipedon with an argilic B-horizon and usually occurs in a rolly to a gently rolling landscape. PV has moderate drainage compared to
Final considerations and conclusions
Organic matter, total iron, silt, sand and minerals, such as quartz, magnetite, kaolinite and smectite, were the most important characteristics that influenced the SR intensity and spectral features, allowing for characterization and discrimination of soils. High absorption features centered at 1900 nm are related to OH molecules in free water (mostly adsorbed) present in 2:1 minerals. Descriptive analyses were helpful when analysed over the whole spectrum, including absorption features and
Acknowledgments
We acknowledge the São Paulo Financial Support Foundation (FAPESP) for the spectroradiometer IRIS (p. no 95/6259-6) and the work (p. no 95/9641-9) and the National Brazilian Research (CNPq) for the first author scholarship (p. no 300371/96-9).
References (48)
- et al.
Variability of laboratory measured soil lines of soil from southeastern Brazil
Remote Sens. Environ.
(1998) - et al.
A modified hapke model for soil bi-directional reflectance
Remote Sens. Environ.
(1996) - et al.
Opportunities and limitations for image-based remote sensing in precision crop management
Remote Sens. Environ.
(1997) - et al.
Mapping the iron oxide content of dune sands, Namib Sand Sea, Namı́bia, using Landsat Thematic Mapper Data
Remote Sens. Environ.
(1997) Soil databases
- et al.
Near infrared analysis (NIRA) as a method to simultaneously evaluate spectral featureless constituents in soils
Soil Sci.
(1995) - et al.
Soil reflectance
Soils and the Greenhouse Effect
(1990)- et al.
Reflection of radiant energy from soils
Soil Sci. Soc. Am. J.
(1965) - Camargo, A.O., Moniz, de A.C., Jorge, J.A., Valadares, J.M., 1986. Métodos de análise quı́mica, mineralógica e fı́sica...
Classificação de solos usada em levantamentos pedológicos no Brasil
Soc. Bras. Ciênc. Solo
A pedologia direcionada ao manejo de solos
Bol. Inf. Soc. Bras. Ci. Solo Viçosa
Characterization and discrimination of soils by their reflected electromagnetic energy
Pesqui. Agropecu.
Detecção de solos erodidos pela avaliação de dados espectrais
Rev. Bras. Ciênc. Solo (Braz. J. Soil Sci.)
Alteration of soil properties through a weathering sequence as evaluated by spectral reflectance
Soil Sci. Soc. Am. J.
Soil characterization and discrimination on a toposequence over diabase through their reflected energy
Cronologia do uso da terra na microbacia hidrográfica do Ceveiro, em Piracicaba, SP
Pesqui. Agropecu. Bras.
Comportamento espectral (450–2.450 nm) de solos tropicais de São Paulo
Rev. Bras. Ciên. Solo
Particle-size analysis
Cited by (185)
A soil spectral library of New Zealand
2023, Geoderma RegionalSoil order knowledge as a driver in soil properties estimation from Vis-NIR spectral data – Case study from northern Karnataka (India)
2023, Geoderma RegionalCitation Excerpt :Visible and near-infrared (Vis-NIR, 350–2500 nm) laboratory spectroscopy provides a complementary method to wet chemistry methods for estimating soil properties (e.g., Viscarra Rossel et al., 2006; Demattê et al., 2004; Stenberg et al., 2010; McBride, 2022) and is non-destructive, rapid, low-cost, efficient, repeatable and reproducible with an acceptable degree of accuracy.