Remote Sensing from Ground to Space Platforms Associated with Terrain Attributes as a Hybrid Strategy on the Development of a Pedological Map

Abstract

There is a consensus about the necessity to achieve a quick soil spatial information with few human resources. Remote/proximal sensing and pedotransference are methods that can be integrated into this approach. On the other hand, there is still a lack of strategies indicating on how to put this in practice, especially in the tropics. Thus, the objective of this work was to suggest a strategy for the spatial prediction of soil classes by using soil spectroscopy from ground laboratory spectra to space images platform, as associated with terrain attributes and spectral libraries. The study area is located in São Paulo State, Brazil, which was covered by a regular grid (one per ha), with 473 boreholes collected at top and undersurface. All soil samples were analyzed in laboratory (granulometry and chemical), and scanned with a VIS-NIR-SWIR (400-2500 nm) spectroradiometer. We developed two traditional pedological maps with different detail levels for comparison: TFS-1 regarding orders and subgroups; and TFS-2 with additional information such as color, iron and fertility. Afterwards, we performed a digital soil map, generated by models, which used the following information: (i) predicted soil attributes from undersurface layer (diagnostic horizon), obtained by using a local spectral library; (ii) spectral reflectance of a bare soil surface obtained by Landsat image; and (iii) derivative of terrain attributes. Thus, the digital map was generated by a combination of three variables: remote sensing (Landsat data), proximal sensing (laboratory spectroscopy) and relief. Landsat image with bare soil was used as a first observation of surface. This strategy assisted on the location of topossequences to achieve soil variation in the area. Afterwards, spectral undersurface information from these locations was used to modelize soil attributes quantification (156 samples). The model was used to quantify samples in the entire area by spectra (other 317 samples). Since the surface was bare soil, it was sampled by image spectroscopy. Indeed, topsoil spectral laboratory information presented great similarity with satellite spectra. We observed angle variation on spectra from clayey to sandy soils as differentiated by intensity. Soil lines between bands 3/4 and 5/7 were helpful on the link between laboratory and satellite data. The spectral models of soil attributes (i.e., clay, sand, and iron) presented a high predictive performance (R 2 0.71 to 0.90) with low error. The spatial prediction of these attributes also presented a high performance (validations with R 2 > 0.78). The models increased spatial resolution of soil weathering information using a known spectral library. Elevation (altitude) improved mapping due to correlation with soil attributes (i.e., clay, iron and chemistry). We observed a close relationship between soil weathering index map and laboratory spectra + image spectra + relief parameters. The color composite of the 5R, 4G and 3B had great performance on the detection of soils along topossequences, since...

Surface Spectroscopy of Oxisols, Entisols and Inceptisol and Relationships with Selected Soil Properties

“…This was possible because the soil surface patterns are related to soil class-specific variations in the underlying horizons and throughout the soil profile. Thus, it is possible to relate surface spectral curves with soil classes, as demonstrated by Demattê et al (2007Demattê et al ( , 2009Demattê et al ( , 2016 and Nanni et al (2011Nanni et al ( , 2012. However, these relationships should be evaluated with caution, because in some cases, the soil surface information alone is insufficient for soil classification.…”

Digital Soil Mapping Using Machine Learning Algorithms in a Tropical Mountainous Area

“…Up to then, studies had not used open source software, such as SAGA and R. In the following years, due to the development and free sharing of codes for soil modeling on open source platforms, several models run in R software became popular in research dealing with soil mapping. After 2011, Logistic regression continued to feature among the most used models for mapping soil classes in Brazil (Souza, 2013;Vasques et al, 2015;Chagas et al, 2017;Jeune et al, 2018), along with the following: ANNs (Arruda et al, 2016), maximum likelihood (Demattê et al, 2016), fuzzy logic and expert knowledge (Silva et al, 2014), and tree-based models (Bazaglia Filho et al, 2013;Höfig et al, 2014;Rizzo et al, 2016).…”

Mapeamento pedológico digital via regressão geograficamente ponderada e lógica booleana: uma estratégia integrada entre dados espectrais terrestres e de satélite

“…Nessas estratégias, estimam-se teores de atributos ou classes de solos em locais desconhecidos a partir amostras conhecidas que têm correlação com dados espectrais de satélite e/ou dados de relevo, via técnicas multivariadas, e assim obter mapas com classes ou atributos do solo. Observa-se que os trabalhos realizados nos estudos de predição se dividem entre estimativas de atributos (Moore et al, 1993;Odeh et al, 2006;Junior et al, 2008;Dematte et al, 2009;Mulder et al, 2011;Gerighausen et al, 2012;Genú et al, 2013;Franceschini et al, 2015) e estimativa de classes de solos (Demattê et al, 2004(Demattê et al, , 2016Ziadat, 2007;Figueiredo et al, 2008;ten Caten et al, 2013ten Caten et al, , 2009Genú e Demattê, 2012;Arruda et al, 2013;Giasson et al, 2013;Adhikari et al, 2014;Fiorio et al, 2014;Teske et al, 2015;Dias et al, 2015) via imagens de satélite e/ou dados de relevo. Além das estratégias de predição via imagens de satélite e/ou dados de relevo, também se apresenta como importante ferramenta, a biblioteca espectral.…”

“…Validação da predição da reflectância de subsuperfície a partir da reflectância imagem de satélite de superfície via métodos estatísticos multivariados Argila espacializada pelo método GWR a partir das bandas TM MSEKG; 3 Argila espacializada pelo método MRL a partir das bandas TM MSEKG; 4 Bandas TM MSEKGU; 5 Argila espacializada pelo método GWR a partir das bandas tm MSEKGU; 6 Argila espacializada pelo método MRL a partir das bandas TM MSEKGU; MRL: Regressão Linear Multipla; GWR: Regressão Geograficamente Ponderada; MSEKG: Mosaico de solo exposto Krigado; MSEKGU: Mosaico de solo exposto unido com o mosaico de solo exposto Krigado; R²: Coeficiente de determinação a 0.05 de significância; RMSE: Raiz Quadrada do Erro Médio Quadrático; RPIQ: Ratio of Performance/ Prediction to Interquartile.Os melhores resultados das variáveis independentes de origem do MSEKG em relação MSEKGU, podem ocorrer pelas transições de reflectâncias não suaves presentes no MSEKGU (Figura 6). Se observado em uma composição R:5, G:4 e B:3, as colorações em tons roxos mais escuros tem relação direta, por exemplo, com maiores teores de argila, enquanto na coloração em tons rosa mais claros pode indicar maiores teores de areia(Demattê et al, 2016). Assim, essas transições mais abruptas observadas no mosaico MSEKGU (Figura 6), podem ter influência negativa na qualidade das predições.…”

“…Transições essas, que são suavizadas pelo efeito da krigagem na formação do mosaico MSEKG, onde valores de máximo da distribuição estatística dos dados tendem a ser subestimados e valores de mínimo superestimados(Da Rocha et al, 2007). ;Araújo et al (2015);Demattê et al (2016). A argila estimada pelo método MRL a partir do MSEKGU, mostrou resultado não confiável (R²<0.5).…”