Alexandre Cândido Xavier

Multi‐temporal analysis of MODIS data to classify sugarcane crop

This paper presents a feasibility study using multi‐temporal Enhanced Vegetation Index (EVI) from Moderate Resolution Imaging Spectroradiometer (MODIS) data to classify sugarcane crop. This study was carried out in São Paulo State which is the major sugarcane producer in Brazil, occupying more than 3.1 million hectares. Cloud‐free MODIS images (16 days mosaics) were acquired over a period of almost 15 months. Samples of sugarcane and non‐sugarcane were randomly selected and cluster analysis was performed to establish similar EVI temporal behaviour clusters. It was observed that EVI was sensitive to variations in land‐use cover mainly due to phenology and land management practices. Therefore, selection of sugarcane samples with similar EVI temporal behaviour for supervised classification was difficult due to both large planting and large harvesting periods. Consequently, cluster analysis was chosen to carry out an unsupervised classification. The best results were obtained in regions occupied by: natural and planted forest, soybean, peanuts, water bodies and urban areas which contrasted with the temporal‐spectral behaviour of sugarcane. The lowest performance was observed mainly in regions dominated by pasture, which has similar temporal‐spectral behaviour to sugarcane. This study provided useful information to define a MODIS image classification procedure for sugarcane crop for the whole State area based on the large amount of cloud‐free MODIS images when compared with other currently available optical sensors.

Variabilidade espacial de atributos químicos de um Latossolo Vermelho-Amarelo húmico cultivado com café

The objective of this study was to analyze the spatial variability of chemical attributes of an Oxisol. The experiment was conducted with Coffea arabica L., variety Catuai, at a depth of 0-0.2 m in a grid area totaling 50 points. The studied chemical attributes were: P, K, Ca, Mg, Na, S, Al, pH, H + Al, SB, t, T, V, m, MO, ISNa, equilibrium P and micronutrients (Zn, Fe, Mn, Cu, and B). The data were analyzed by descriptive statistical analysis, followed by principal component with subsequent cluster analysis. The existence of spatial dependence was verified and the degree quantified using the principal components and geostatistics. The method of multivariate analysis based on principal components provided interpretable components; the first were correlated with acidity and the second with soil organic matter. Techniques of multivariate analysis, in combination with geostatistics, facilitated the evaluation of soil variability. The principal components 1 and 2 indicated moderate spatial dependence, with greater spatial continuity in component 1, which allowed a better characterization of soil acidity. Reduced availability of soil nutrients and considerable acidity were observed along the slope, lower levels of organic matter in the center of the area and high Al contents in the upper part.

Hyperspectral field reflectance measurements to estimate wheat grain yield and plant height

Hyperspectral crop reflectance data are useful for several remote sensing applications in agriculture, but there is still a need for studies to define optimal wavebands to estimate crop biophysical parameters. The objective of this work is to analyze the use of narrow and broad band vegetation indices (VI) derived from hyperspectral field reflectance measurements to estimate wheat (Triticum aestivum L.) grain yield and plant height. A field study was conducted during the winter growing season of 2003 in Campinas, Sao Paulo State, Brazil. Field canopy reflectance measurements were acquired at six wheat growth stages over 80 plots with four wheat cultivars (IAC-362, IAC-364, IAC-370, and IAC-373), five levels of nitrogen fertilizer (0, 30, 60, 90, and 120 kg of N ha-1) and four replicates. The following VI were analyzed: a) hyperspectral or narrow-band VI (1. optimum multiple narrow-band reflectance, OMNBR; 2. narrow-band normalized difference vegetation index, NB_NDVI; 3. first- and second-order derivative of reflectance; and 4. four derivative green vegetation index); and b) broad band VI (simple ratio, SR; normalized difference vegetation index, NDVI; and soil-adjusted vegetation index, SAVI). Hyperspectral indices provided an overall better estimate of biophysical variables when compared to broad band VI. The OMNBR with four bands presented the highest R2 values to estimate both grain yield (R2 = 0.74; Booting and Heading stages) and plant height (R2 = 0.68; Heading stage). Best results to estimate biophysical variables were observed for spectral measurements acquired between Tillering II and Heading stages.

Evaluation of the performance of the different methods of interpolaters for parameters of the climatologic water balance

A geoestatistica esta associada a uma classe de tecnicas utilizadas para analisar e inferir valores de uma variavel distribuida no espaco ou no tempo, mediante o que se propos, no presente trabalho avaliar, atraves de diferentes tecnicas de interpolacao, os seguintes parâmetros climaticos: precipitacao, deficiencia hidrica, excedente hidrico, evapotranspiracao potencial, evapotranspiracao real e disponibilidade hidrica, no estado do Espirito Santo. Para tanto, utilizaram-se dados meteorologicos de temperatura do ar e precipitacao pluviometrica, compreendidos no periodo de 1977 a 2006, para o calculo do balanco hidrico climatologico conforme metodo proposto por Thornthwaite & Mather (1955), adotando uma capacidade de armazenamento de 100 mm. Os resultados mostram que o metodo da krigagem e o mais eficiente para a espacializacao dos parâmetros climaticos, baseado no menor valor da Raiz do Erro Medio Quadratico (REMQ) e outros parâmetros calculados que auxiliaram na escolha do melhor modelo.

Monitoring leaf area index at watershed level through NDVI from Landsat-7/ETM+ data

O Indice de Area Foliar (IAF) e um importante parâmetro do dossel da vegetacao, sendo utilizado, por exemplo, para estimar a evapotranspiracao, um importante componente do ciclo hidrologico. Analisou-se a relacao entre o IAF, medido em campo, e o indice de vegetacao NDVI, de quatro datas (a partir de dados Landsat-7/ETM+) e, com este indice, foram gerados e analisados mapas de IAF da area de estudo para as diferentes datas. O IAF foi coletado mensalmente em campo, com o equipamento LAI-2000, em areas de cana-de-acucar, pastagem, milho, eucalipto e floresta riparia. A relacao entre o IAF e o NDVI foi melhor ajustada por um modelo potencial, sendo que de 57% a 72% da variacao do NDVI puderam ser explicados pelo IAF. Os mapas de IAF gerados por modelos empiricos da relacao entre o IAF e NDVI mostraram razoavel precisao (o erro padrao da estimativa de IAF variou de 0,42 a 0,87 m2 m-2). O valor medio de IAF para cada mapa mensal mostrou-se relacionado com a precipitacao total dos tres meses anteriores.

Performance evaluation of rainfall estimates by TRMM Multi‐satellite Precipitation Analysis 3B42V6 and V7 over Brazil

Time series precipitation data generated by the Tropical Rainfall Measuring Mission (TRMM) have been used as a possible solution for providing rainfall information for ungauged regions. We evaluated the quality of TRMM Multi-satellite Precipitation Analysis (TMPA) Version 6 (3B42V6) and Version 7 (3B42V7) products on a daily and monthly basis for a 14 year time series by comparing with gridded ground-based rainfall data from ~3625 rain gauges distributed throughout Brazil. The results show that daily estimates are inaccurate for both Versions 6 and 7 (the refined index of agreement, dr, was less than 0.6 in most of the analyzed pixels). In general, both versions perform well on monthly basis (dr > 0.75), but no significant improvement between them could be identified with the exception of local areas. TMPA performed poorly in the northwest region, where rainfall depths are higher in Brazil; however, the quality of the ground-based data is poor in this region because of low gauge density. Based on a seasonal analysis, we found that TMPA performed better during the dry seasons and that some improvements, although not significant, between successive versions took place over the northeast, southeast, and south regions. This study shows the value of remote sensing precipitation for providing reliable, spatiotemporally continuous precipitation at monthly timescales.

Assessing biophysical variable parameters of bean crop with hyperspectral measurements

Recently high spectral resolution sensors have been developed, which allow new and more advanced applications in agriculture. Motivated by the increasing importance of hyperspectral remote sensing data, the need for research is important to define optimal wavebands to estimate biophysical parameters of crop. The use of narrow band vegetation indices (VI) derived from hyperspectral measurements acquired by a field spectrometer was evaluated to estimate bean (Phaseolus vulgaris L.) grain yield, plant height and leaf area index (LAI). Field canopy reflectance measurements were acquired at six bean growth stages over 48 plots with four water levels (179.5; 256.5; 357.5 and 406.2 mm) and tree nitrogen rates (0; 80 and 160 kg ha-1) and four replicates. The following VI was analyzed: OSNBR (optimum simple narrow-band reflectivity); NB_NDVI (narrow-band normalized difference vegetation index) and NDVI (normalized difference index). The vegetation indices investigated (OSNBR, NB_NDVI and NDVI) were efficient to estimate LAI, plant height and grain yield. During all crop development, the best correlations between biophysical variables and spectral variables were observed on V4 (the third trifoliolate leaves were unfolded in 50 % of plants) and R6 (plants developed first flowers in 50 % of plants) stages, according to the variable analyzed.

Métodos geoestatísticos no estudo da resistência do solo à penetração em trilha de tráfego de tratores na colheita de madeira

This work was carried out to evaluate soil penetration resistance (SPR) in a transversal section of the traffic trail of the tractors Feller-buncher and Skidder during wood harvesting. Geostatistic techniques were used to determine spatial dependence and kriging interpolation. Semivariogram models were fit to determine spatial dependence before traffic and after cutting and skidding. The SPR variation coefficient in the trail reduced with traffic. The result showed a moderate SPR spatial dependence in the soil section before and after cutting,with the spherical model being adjusted. The linear model was adjusted to semivariance after wood skidding, indicating the need to use a larger grid to represent the continuity of the attribute. The geostatistic methodology was found to be a good tool to estimate the values in non-sampled points and in SPR contour map building.

Análise da influência de diferentes tamanhos e composições de amostras no ajuste de uma relação hipsométrica para Eucalyptus grandis

The adjustment of hypsometric relations using data from forest inventory aiming at the estimation of individual tree volumes is a common task in Brazilian forestry companies. This study used a data set of 481 trees of Eucalyptus grandis from a forest inventory. Simulation was carried out varying sample sizes from 6 to 204 sample-trees. For each sample size it was run 10,000 simulations for each size samples selected randomly. This study aimed at evaluating the effect of sample size and structure on the precision of a previously chosen hypsometric model which had, as independent variables, the Diameter at Breast Height (DBH) and the tree age. It was also evaluated the effect of these covariates on the tree heights estimates. It was concluded that a sample with at least 27 trees could more precisely estimate tree height when compared to smaller samples. In addition, sample sizes containing more than 27 trees did not shown significant improving on the precision of the height estimates. For the hipysometric model, the variable DBH was more valuable to estimate the tree height than the covariate age. For the trees with larger DBH, the variable age showed to be more influent.

Comparação entre métodos de amostragem do solo para recomendação de calagem e adubação do cafeeiro conilon

This study aimed to evaluate the soil sampling conventional methodology with spatial analysis for liming and fertilization with nitrogen, phosphorus and potassium on soil cultivated with conilon coffee. The trial was carried out during the 2004/2005 - 2005/2006 harvests in a 1.0 ha area with samples collected at 0-0.20 m depth. Fifteen sub samples in zigzag were collected for the conventional method forming a compound sample; as for the spatial method, 109 georeferenced points formed a sample grid. After the analysis results, the liming and fertilization needs were calculated based on the function of the element content in the soil and on the plant expected yield. Data were analyzed by both the classical statistics (descriptive and exploratory) and spatial analysis, using geostatistics techniques (modeling of the spatial variability structure and inferences) and geoprocessing (map algebra). Except for phosphorus in 2005 and potassium in 2006, every other recommendation showed spatial dependence. Data analysis by the spatial method provided the identification of zones with deficient or excessive liming and fertilization which could not be defined by the conventional sampling method (zigzag).