A flexible spatiotemporal method for fusing satellite images with different resolutions

Satellite data for studying surface dynamics in heterogeneous landscapes are missing due to frequent cloud contamination, low temporal resolution, and technological difficulties in developing satellites. A modified spatiotemporal fusion algorithm for predicting the reflectance of paddy rice is presented in this paper. The algorithm uses phenological information extracted from a moderate-resolution imaging spectroradiometer enhanced vegetation index time series to improve the enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM). The algorithm is tested with satellite data on Yueyang City, China. The main contribution of the modified algorithm is the selection of similar neighborhood pixels by using phenological information to improve accuracy. Results show that the modified algorithm performs better than ESTARFM in visual inspection and quantitative metrics, especially for paddy rice. This modified algorithm provides not only new ideas for the improvement of spatiotemporal data fusion method, but also technical support for the generation of remote sensing data with high spatial and temporal resolution.

Section: Calculation Of the Prediction Reflectancementioning

confidence: 95%

Section: Introductionmentioning

confidence: 96%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Modified Spatiotemporal Fusion Algorithm Using Phenological Information for Predicting Reflectance of Paddy Rice in Southern China

Liu

et al. 2018

“…In addition to the spectral mixture analysis in reducing the mixed pixel problem, another approach is the use of data fusion of multi-resolution/sensor data [48][49][50][51]. However, in a large area, the data fusion of Landsat TM and MODIS data may not be cost effective or may be not necessary because Landsat TM image can reliably provide cropland classification.…”

Section: Discussionmentioning

confidence: 99%

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

Zhu

Victoria

et al. 2015

Mapping cropland distribution over large areas has attracted great attention in recent years, however, traditional pixel-based classification approaches produce high uncertainty in cropland area statistics. This study proposes a new approach to map fractional cropland distribution in Mato Grosso, Brazil using time series MODIS enhanced vegetation index (EVI) and Landsat Thematic Mapper (TM) data. The major steps include: (1) remove noise and clouds/shadows contamination using the Savizky-Gloay filter and temporal resampling algorithm based on the time series MODIS EVI data; (2) identify the best periods to extract croplands through crop phenology analysis; (3) develop a seasonal dynamic index (SDI) from the time series MODIS EVI data based on three key stages: sowing, growing, and harvest; and (4) develop a regression model to estimate cropland fraction based on the relationship between SDI and Landsat-derived fractional cropland data. The root mean squared error of 0.14 was obtained based on the analysis of randomly selected 500 sample plots. This research shows that the proposed approach is promising for rapidly mapping fractional cropland distribution in Mato Grosso, Brazil.

“…It is sometimes termed data fusion, or more specifically, pansharpening, when the higher resolution ancillary data is the panchromatic band [35][36][37][38][39]. Hereafter, for consistency, the term downscaling is used throughout this paper.…”

Section: Downscaling Landsat-8 30-m Data To 15 M Using the Panchromatmentioning

confidence: 99%

Landsat 15-m Panchromatic-Assisted Downscaling (LPAD) of the 30-m Reflective Wavelength Bands to Sentinel-2 20-m Resolution

Zhang

Roy

et al. 2017

Abstract:The Landsat 15-m Panchromatic-Assisted Downscaling (LPAD) method to downscale Landsat-8 Operational Land Imager (OLI) 30-m data to Sentinel-2 multi-spectral instrument (MSI) 20-m resolution is presented. The method first downscales the Landsat-8 30-m OLI bands to 15-m using the spatial detail provided by the Landsat-8 15-m panchromatic band and then reprojects and resamples the downscaled 15-m data into registration with Sentinel-2A 20-m data. The LPAD method is demonstrated using pairs of contemporaneous Landsat-8 OLI and Sentinel-2A MSI images sensed less than 19 min apart over diverse geographic environments. The LPAD method is shown to introduce less spectral and spatial distortion and to provide visually more coherent data than conventional bilinear and cubic convolution resampled 20-m Landsat OLI data. In addition, results for a pair of Landsat-8 and Sentinel-2A images sensed one day apart suggest that image fusion should be undertaken with caution when the images are acquired under different atmospheric conditions. The LPAD source code is available at GitHub for public use.