Mapping Fractional Cropland Distribution in Mato Grosso, Brazil Using Time Series MODIS Enhanced Vegetation Index and Landsat Thematic Mapper Data

Zhu, Changming; Lu, Dengsheng; Victoria, D. de C.; Dutra, Luciano Vieira

doi:10.3390/rs8010022

Cited by 31 publications

(25 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The identification of irrigation pixels is consistent with actual irrigation samples in field surveys and statistical areas ( Table 3). It is possible to supervise the distribution and the area of irrigation automatically using satellite-based datasets, avoiding visual interpretation [55][56][57] of high-resolution images [58] and much fieldwork, as well as substantially improving the method's repeatability and applicability.…”

Section: Discussionmentioning

confidence: 99%

Mapping Irrigated Areas of Northeast China in Comparison to Natural Vegetation

Xiang

Liu

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

Accurate information about the location and extent of irrigation is fundamental to many aspects of food security and water resource management. This study develops a new method for identifying irrigation in northeastern China by comparing canopy moisture between the cropland and adjacent natural ecosystems (i.e., forests). This method is based on two basic assumptions, which we validated using field survey data. First, the canopy moisture of irrigated cropland, indicated by a satellite-based land surface water index (LSWI), is higher than that of the adjacent forest. Second, the difference in LSWI between irrigation cropland and forest is larger in arid regions than in humid regions. Based on the field survey and statistical dataset, our method performed well in indicating spatial variations of irrigated areas. Results from this study suggest that our method is a promising tool for mapping irrigated areas, as it is a general and repeatable method that does not rely on training samples and can be applied to other regions.

show abstract

Section: Discussionmentioning

confidence: 99%

Mapping Irrigated Areas of Northeast China in Comparison to Natural Vegetation

Xiang

Liu

et al. 2019

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…The key identification stage selection is fundamental for constructing a seasonal dynamic index, which is critical for cropland mapping in the Brazilian Amazon region. As almost all images are contaminated by clouds in the rainy season, a feasible means is to use slices of discrete time series data instead of entire continuous time series data for a year for crop mapping [26] . As the EVI profiles in cropland area vary regularly at different stages (e.g., the sowing, growing, and harvest stage), three key identification stages: the sowing (Stage 1, DOY:225-289), growing (Stage 2, DOY:305-001), and harvest (Stage 3, DOY:017-081) seasons are the key identification stages for cropland mapping [26] .…”

Section: Key Identification Stage and Seasonal Dynamic Indexmentioning

confidence: 99%

“…The seasonal dynamic index (SDI) model was proposed by Zhu et al. The model can be elaborated by Equations (2)-(8) [26] : (8) where, SDI represents the seasonal dynamic index; and SDI 1 and SDI 2 correspond to the seasonal dynamic index at different stages; EVI d , EVI g , and EVI h are cloud-free EVI composites from the dry to wet season transition, the growth, and the harvest season, respectively; EVI 225 , EVI 241 , …, EVI 353 , are the multi-temporal MODIS EVI products, and the number subscript is the acquired day of the year (DOY); Slp mask is the topographic factor mask where the slope is derived from STRM data; Slp mask is a slope mask;…”

Section: Key Identification Stage and Seasonal Dynamic Indexmentioning

confidence: 99%

“…It has been noticed that previous methods primarily relied on pixel level "hard" classifications. These methods are able to identify cropland qualitatively; however , they have the disadvantage of not being able to extract cropland fractions accurately and quantitatively because the mixed pixel problem is unavoidable in coarse spatial resolution imagery (e.g., MODIS), resulting in poor area estimation and inaccurate s patial patterns [26] . Therefore, sub-pixel level "soft" classification and fractional cropland information extraction has attracted more attention [22][23][24][25][26] .…”

Section: Introduction mentioning

confidence: 99%

See 1 more Smart Citation

Mapping fractional cropland covers in Brazil through integrating LSMA and SDI techniques applied to MODIS imagery

Zhu

Zhang

Huang

2019

International Journal of Agricultural and Biological Engineering

Self Cite

View full text Add to dashboard Cite

MODIS time-series imagery is promising for generating regional and global land cover products. For Brazil, however, accurate fractional cropland covers (FCC) information is difficult to obtain due to frequent cloud coverage and the mixing-pixel problem. To address these problems, this study developed an innovative approach to mapping the FCC of the Mato Grosso State, Brazil through integrating Linear Spectral Mixture Analysis (LSMA) and Seasonal Dynamic Index (SDI) models. With MOD13Q1 time-series EVI imagery, a SDI was developed to represent the phenology of croplands. Furthermore, fractional land covers (e.g., vegetation, soil, and low albedo components) were derived with the LSMA algorithms . A stepwise regression model was established to estimate the FCC at the regional scale . Finally, ground truth cropland cover information was extracted from Landsat TM imagery using a hybrid method. Results indicated that the combination of multiple feature variables produced better results when compared with individual variables. Through cross-validation and comparative analysis, the coefficient of determination (R 2 ) between the reference and estimated FCCs reached 0.84 with a Root Mean Square Error (RMSE) of 0.13. This indicates that the proposed method effectively improved the accuracy of fractional cropland mapping. When compared to the traditional per-pixel "hard" classification, the sub-pixel level maps illustrated detailed cropland spatial distribution patterns. Zhu C M, et al. Mapping FCC in Brazil through integrating LSMA and SDI techniques applied to MODIS imagery Vol.

show abstract

“…Currently, the Moderate-Resolution Imaging Spectroradiometer (MODIS) sensor on board the Terra satellite provides an adequate imaging configuration for crop monitoring based on (1) an almost-daily revisit time; (2) a moderate spatial resolution of 250-m, considered adequate for mapping large-scale agricultural fields (Lobell and Asner 2004); and (3) a geometric quality that is high enough for time series analysis (Justice et al 2002). A study conducted by Lobell and Asner (2004) concluded that MODIS images have considerable advantages in the characterization of extensive agricultural crops, mainly due to their higher temporal resolution, and many studies focused on Brazil have highlighted the efficiency of the MODIS time series of vegetation indices for mapping cropland, crop expansion Macedo et al 2006;Gusso et al 2014, Zhu et al 2016 and cropping system management (Galford et al 2008;Arvor et al 2012;Brown et al 2013).…”

Section: Introductionmentioning

confidence: 99%

Analysis of soybean cropland expansion in the southern Brazilian Amazon and its relation to economic drivers

2017

View full text Add to dashboard Cite

The agricultural dynamics of soybean expansion have long been recognized as a major driver of excessive land cover change on the southwestern border of the Brazilian Amazon. The hypothesis that the soybean market exerts an influence on land use was investigated by the association between economic indicators and soybean crop dynamics in the state of Mato Grosso (western Brazil). We integrated a historical series of satellite data of soybean cropland expansion and the two main economic variables (selling prices and production costs) for soybean in Mato Grosso. We focused on the relation between profit (the difference between the average soybean price and production costs) and land-use transition to soybean from 2001 to 2013. The spatially explicit analysis showed that the overall accuracy between the resulting first-time use and the most recent soybean crop area in 2013 was 96.75%, with a Kappa index of 0.63. However, dissimilar values found between Omission and Commission indicators suggest that most of the expanded areas prior to 2013 (5.57 million ha) were under a high dynamical range of land uses. Although there is no direct relation between either the deforestation rate or expansion trends (first-time-use rate) and profit, the results strongly suggest (R 2 =0.81) that profit exerts a direct and non-negligible influence on the evolution of consolidated land use for soybean in Mato Grosso State. KEYWORDS: land use, agriculture, soybean mapping, Amazon biome Análise da expansão do cultivo da soja no sul da Amazônia brasileira utilizando imagens de satélite e sua relação com forçantes econômicas RESUMO A dinâmica agrícola relacionada à expansão da soja tem sido reconhecida como um dos principais fatores da excessiva conversão da cobertura do solo no Estado de Mato Grosso, no sudoeste da Amazônia brasileira. A hipótese de que o mercado de soja exerce influência no uso do solo, foi investigada pela associação de indicadores econômicos com a dinâmica da cultura de soja no Mato Grosso. Integramos séries históricas de dados de satélite para expansão da área de cultivo de soja e de suas duas principais variáveis econômicas associadas (preço de venda e custo de produção). Enfocamos a relação entre lucro (a diferença entre as médias do preço da soja e do custo de produção) e a conversão do uso do solo para soja de 2001 a 2013. A análise espacial explícita revelou que a precisão global na comparação entre o mapa resultante de first-time-use de cultivo de soja em 2013 foi de 96,75%, com índice Kappa de 0,63. Entretanto, a divergência obtida entre os indicadores de erro por comissão e omissão, sugerem que a maior parte da expansão da soja ocorrida antes de 2013 (5,57 milhões de ha) esteve sob influência de uma intensa dinâmica de uso do solo. Embora seja claro não haver relação direta entre a taxa de desmatamento ou tendências de expansão (taxa de uso pela primeira vez) e o lucro, os resultados sugerem fortemente (R 2 =0,81) que o lucro exerce influência direta e não-negligenciável na evolução do uso do solo consoli...

show abstract

Mapping Fractional Cropland Distribution in Mato Grosso, Brazil Using Time Series MODIS Enhanced Vegetation Index and Landsat Thematic Mapper Data

Cited by 31 publications

References 51 publications

Mapping Irrigated Areas of Northeast China in Comparison to Natural Vegetation

Mapping Irrigated Areas of Northeast China in Comparison to Natural Vegetation

Mapping fractional cropland covers in Brazil through integrating LSMA and SDI techniques applied to MODIS imagery

Analysis of soybean cropland expansion in the southern Brazilian Amazon and its relation to economic drivers

Contact Info

Product

Resources

About