USGS Science⌕ Search
Home/Archive/USGS

USGS · 70035284

Large area scene selection interface (LASSI): Methodology of selecting landsat imagery for The Global Land Survey 2005

S. Franks·J. G. Masek·R.M.K. Headley·J. Gasch·T. Arvidson
View original source ↗Export metadata ↓

Abstract

The Global Land Survey (GLS) 2005 is a cloud-free, orthorec-tified collection of Landsat imagery acquired during the 2004 to 2007 epoch intended to support global land-cover and ecological monitoring. Due to the numerous complexities in selecting imagery for the GLS2005, NASA and the U.S. Geological Survey (USGS) sponsored the development of an automated scene selection tool, the Large Area Scene Selection Interface (LASSI), to aid in the selection of imagery for this data set. This innovative approach to scene selection applied a user-defined weighting system to various scene parameters: image cloud cover, image vegetation greenness, choice of sensor, and the ability of the Landsat-7 Scan Line Corrector (SLC)-off pair to completely fill image gaps, among others. The parameters considered in scene selection were weighted according to their relative importance to the data set, along with the algorithm’s sensitivity to that weight. This paper describes the methodology and analysis that established the parameter weighting strategy, as well as the post-screening processes used in selecting the optimal data set for GLS2005.

Explore related subjects

Photogrammetric Engineering and Remote SensingArticle

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

S. Franks, J. G. Masek, R.M.K. Headley, J. Gasch, T. Arvidson. 2009. Large area scene selection interface (LASSI): Methodology of selecting landsat imagery for The Global Land Survey 2005. https://doi.org/10.14358/pers.75.11.1287

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2025

Evaluating Three-Dimensional Elevation Program lidar consistency and accuracy at scale using cloud-native, open-source methods

The U.S. Geological Survey three-dimensional elevation program (3DEP) has significantly expanded national lidar coverage, necessitating scalable, reproducible methods for assessing data quality across diverse terrains and acquisition conditions. This study introduces a cloud-native, open-source workflow designed to evaluate the geometric accuracy and consistency of 3DEP lidar data sets at a national scale. Leveraging tools such as the Point Data Abstraction Library, Open3D, and Amazon Web Services infrastructure, the workflow integrates global navigation satellite system‐surveyed ground control points and terrestrial laser scanning data to validate airborne lidar collections. Two case studies demonstrate the application of this process. In Puerto Rico, the process identified vertical biases and inconsistencies in vegetated areas, while in Iowa and Arizona, the process confirmed high vertical accuracy with minimal bias. The results underscore the effectiveness of combining cloud computing with open-source tools to perform large-scale lidar data quality assessments. This process offers a reproducible, efficient solution for nationwide validation of 3DEP data sets, supporting enhanced decision-making in geospatial applications.

Photogrammetric Engineering and Remote Sensing↗
USGS2025

Artificial neural network multilayer perceptron models to classify California’s crops using Harmonized Landsat Sentinel (HLS) data

Advances in remote sensing and machine learning are enhancing cropland classification, vital for global food and water security. We used multispectral Harmonized Landsat 8 Sentinel-2 (HLS) 30-m data in an artificial neural network (ANN) multi-layer perceptron (MLP) model to classify five crop classes (cotton, alfalfa, tree crops, grapes, and others) in California's Central Valley. The ANN MLP model, trained on 2021 data from the United States Department of Agriculture's Cropland Data Layer, was validated by classifying crops for an independent year, 2022. Across the five crop classes, the overall accuracy was 74%. Producer's and user's accuracies ranged from 65% to 87%, with cotton achieving the highest accuracies. The study highlights the potential of using deep learning with HLS time series data for accurate global crop classification.

California↗