A global analysis of the temporal availability of PlanetScope high spatial resolution multi-spectral imagery

Roy, David P.; Huang, Haiyan; Houborg, Rasmus; Martins, Vitor Souza

doi:10.1016/j.rse.2021.112586

Cited by 106 publications

(55 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The image processing component was designed for PlanetScope Analytic surface reflectance imagery (PlanetTeam, 2018), which provides three visual (red, green, blue) and near-infrared bands at 3.7 m resolution at nominal daily frequency. The images are provided as ortho-rectified and converted to surface reflectance, although there are residual errors from inter-sensor differences and the radiometric normalization process (Houborg and McCabe, 2018), variation in the orientation of scene footprints, as well as a high frequency of cloud cover over the study region (Wilson and Jetz, 2016;Roy et al, 2021) that are not fully captured by the provided cloud masks. To minimize the effect of these residual errors, we developed a procedure for creating temporal composites of the primary growing and nongrowing seasons within a single 12-month period.…”

Section: Image Compositingmentioning

confidence: 99%

High Resolution, Annual Maps of Field Boundaries for Smallholder-Dominated Croplands at National Scales

Estes

Song

et al. 2022

Front. Artif. Intell.

View full text Add to dashboard Cite

Mapping the characteristics of Africa’s smallholder-dominated croplands, including the sizes and numbers of fields, can provide critical insights into food security and a range of other socioeconomic and environmental concerns. However, accurately mapping these systems is difficult because there is 1) a spatial and temporal mismatch between satellite sensors and smallholder fields, and 2) a lack of high-quality labels needed to train and assess machine learning classifiers. We developed an approach designed to address these two problems, and used it to map Ghana’s croplands. To overcome the spatio-temporal mismatch, we converted daily, high resolution imagery into two cloud-free composites (the primary growing season and subsequent dry season) covering the 2018 agricultural year, providing a seasonal contrast that helps to improve classification accuracy. To address the problem of label availability, we created a platform that rigorously assesses and minimizes label error, and used it to iteratively train a Random Forests classifier with active learning, which identifies the most informative training sample based on prediction uncertainty. Minimizing label errors improved model F1 scores by up to 25%. Active learning increased F1 scores by an average of 9.1% between first and last training iterations, and 2.3% more than models trained with randomly selected labels. We used the resulting 3.7 m map of cropland probabilities within a segmentation algorithm to delineate crop field boundaries. Using an independent map reference sample (n = 1,207), we found that the cropland probability and field boundary maps had respective overall accuracies of 88 and 86.7%, user’s accuracies for the cropland class of 61.2 and 78.9%, and producer’s accuracies of 67.3 and 58.2%. An unbiased area estimate calculated from the map reference sample indicates that cropland covers 17.1% (15.4–18.9%) of Ghana. Using the most accurate validation labels to correct for biases in the segmented field boundaries map, we estimated that the average size and total number of field in Ghana are 1.73 ha and 1,662,281, respectively. Our results demonstrate an adaptable and transferable approach for developing annual, country-scale maps of crop field boundaries, with several features that effectively mitigate the errors inherent in remote sensing of smallholder-dominated agriculture.

show abstract

Section: Image Compositingmentioning

confidence: 99%

High Resolution, Annual Maps of Field Boundaries for Smallholder-Dominated Croplands at National Scales

Estes

Song

et al. 2022

Front. Artif. Intell.

View full text Add to dashboard Cite

show abstract

“…Analyses of the interplay of landscapes/seascapes and soundscapes is at the forefront of many applications of acoustic remote sensing and FRS is especially interested in advancing this area of research. This research could also involve the integration of acoustic remote sensing data with that from other remote sensing platforms, such as those from LiDAR (e.g., Asner et al, 2012), hyperspectral (e.g., Asner and Martin, 2009) and multispectral imagery (e.g., Roy et al, 2021;Yan and Roy, 2021).…”

Section: Advances In Seascape Ecologymentioning

confidence: 99%

Grand Challenges in Acoustic Remote Sensing: Discoveries to Support a Better Understanding of Our Changing Planet

Pijanowski

Brown

2022

Front. Remote Sens.

View full text Add to dashboard Cite

model of the seafloor derived from multibeam bathymetric measurements. Inset image shows how MBES data sets are collected. Imagery courtesy of the Dalhousie University, Seascape Ecology and Mapping Lab (www.seafloormapping.ca); (B) Synthetic Aperture Sonra data set showing a ship wreck and seafloor sediment bedforms. Data is 3 cm horizontal resolution and collected from a towed SAS system shown inset (Kraken Robotics Katfish). Imagery courtesy of Kraken Robotics (https://krakenrobotics.com/); (C) Sub-bottom profiler data showing sediment stratigraphy beneath the seafloor. The profile was collected using a towed sensor platform (Geoforce DTS). Imagery courtesy of Geoforce Group Ltd. (https://www.geoforcegroup.com/).

show abstract

“…The near-daily revisit frequency of CubeSats (Roy et al, 2021) allowed us to track canopy dynamics in detail. For the study period in each site, overall, CubeSats provided 49 -56% of daily data when gapfilled data is removed, which is a higher daily retrieval rate compared to HLS (22 -28%) (Figure 4; Table A5).…”

Section: Flux Tower Footprint Matching and Effects Of Spatial And Temporal Resolution On Gpp Estimationmentioning

confidence: 99%

Matching high resolution satellite data and flux tower footprints improves their agreement in photosynthesis estimates

Kong¹,

Ryu²,

Liu³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Mapping canopy photosynthesis in both high spatial and temporal resolution is essential for carbon cycle monitoring in heterogeneous areas. However, well established satellites in sun-synchronous orbits such as Sentinel-2, Landsat and MODIS can only provide either high spatial or high temporal resolution but not both. Recently established CubeSat satellite constellations have created an opportunity to overcome this resolution trade-off. In particular, Planet Fusion allows full utilization of the CubeSat data resolution and coverage while maintaining high radiometric quality. In this study, we used the Planet Fusion surface reflectance product to calculate daily, 3-m resolution, gap-free maps of the near-infrared radiation reflected from vegetation (NIRvP). We then evaluated the performance of these NIRvP maps for estimating canopy photosynthesis by comparing with data from a flux tower network in Sacramento-San Joaquin Delta, California, USA. Overall, NIRvP maps captured temporal variations in canopy photosynthesis of individual sites, despite changes in water extent in the wetlands and frequent mowing in the crop fields. When combining data from all sites, however, we found that robust agreement between NIRvP maps and canopy photosynthesis could only be achieved when matching NIRvP maps to the flux tower footprints. In this case of matched footprints, NIRvP maps showed considerably better performance than in situ NIRvP in estimating canopy photosynthesis both for daily sum and data around the time of satellite overpass (R 2 = 0.78 vs. 0.60, for maps vs. in situ for the satellite overpass time case). This difference in performance was mostly due to the higher degree of consistency in slopes of NIRvP -canopy photosynthesis relationships across the study sites for flux tower footprint-matched maps. Our results show the importance of matching satellite observations to the flux tower footprint and demonstrate the potential of CubeSat constellation imagery to monitor canopy photosynthesis remotely at high spatio-temporal resolution.

show abstract

A global analysis of the temporal availability of PlanetScope high spatial resolution multi-spectral imagery

Cited by 106 publications

References 62 publications

High Resolution, Annual Maps of Field Boundaries for Smallholder-Dominated Croplands at National Scales

High Resolution, Annual Maps of Field Boundaries for Smallholder-Dominated Croplands at National Scales

Grand Challenges in Acoustic Remote Sensing: Discoveries to Support a Better Understanding of Our Changing Planet

Matching high resolution satellite data and flux tower footprints improves their agreement in photosynthesis estimates

Contact Info

Product

Resources

About