NLCD Readies Improvements for Upcoming Release of 2019 Product Suite

Release Date:

The next edition of the National Land Cover Database (NLCD) will appear more quickly than previous editions, offer accuracy comparable to NLCD 2016, and include previously-unreleased historical information on imperious urban surfaces.

NLCD 2016 Land Cover map of the conterminous United States

National Land Cover Database land cover for the conterminous United States, represented as 16 land cover classes. NLCD is currently working on an update for 2019 land cover conditions. 

NCLD 2019 is scheduled for release during the winter of 2020, marking the most rapid turnaround for a national-scale land cover dataset in the project’s history.

“What we’re trying to do right now for NLCD is kind of unprecedented – doing a 2019 NLCD for release in 2020,” said Terry Sohl, a research physical scientist at the USGS Earth Resources Observation and Science (EROS) Center who recently took on the role of NCLD co-director. “We’re using new and improved methods to decrease the latency between releases.”

NLCD is one of the most widely-used geospatial datasets in the U.S., serving as a basis for understanding the Nation’s landscapes in thousands of studies and applications, trusted by scientists, land managers, students, city planners and many more as a definitive source of U.S. land cover. 

NLCD released its most expansive and detailed database to date in mid-2019, sorting each 30-meter plot of ground in the country into a land cover class – e.g. developed, cropland, wetland – as well as offering information on tree canopy cover, impervious urban surfaces, a roads and energy development layer, and grass/shrub coverage layers for the Western U.S. 

NLCD 2016 depicted conditions across 15 years for the conterminous United States (CONUS), with land cover map layers for seven calendar years: 2001, 2004, 2006, 2008, 2011, 2013, and 2016.

NCLD 2019, in the works for late 2020 or early 2021, will offer the same information for 2019, but will also include several significant improvements. Among them:

- Impervious surface layers (useful for the study of urbanization) that will reflect conditions for the years 2004, 2008, 2013, and 2019,

- A new “impervious descriptor” data layer for each of those years to show the footprint of wind energy in the U.S.,

- Improved accuracy through the integration of decades of forest regrowth data from LANDFIRE and Landsat Analysis Ready Data (ARD) tiles generated for the USGS Land Change Monitoring, Assessment, and Projection (LCMAP) initiative.

The update’s rapid clip and its far-ranging improvements come in large part through the NLCD team’s collaborative partnerships with other EROS land cover mapping projects and other federal agencies.

Color screenshot of Wind Turbine database map

The National Land Cover Databse (NLCD) is working to incorporate wind turbine data, as shown in this screenshot, into its impervious surface products for NLCD 2019. 

“We used every advantage we could in coordinating and collaborating with our partners,” said Jon Dewitz, an EROS-based geographer and NLCD coordinator.

The highlights for NCLD users in the 2019 products are the accuracy improvements through collaboration with other EROS-based land cover projects and the expansive upgrades to impervious surface products through the integration of new data sources, Dewitz said.

The impervious surface maps for 2019 will integrate data on wind energy, oil wells, streets and population density from a variety of sources, each of which will help fine-tune a complex product used for everything from urban heat island studies to stormwater runoff modeling.

“We had some of those (datasets) in 2016, but we have a more complete database for 2019,” Dewitz said. “We’ve incorporated all of those throughout all the years of our impervious products.”

The collaboration factor helped the NLCD team improve its land cover product accuracy, but it’s also sped up the process of creating the Landsat satellite-based map composites that serve as a basis for the production of NLCD thematic and impervious land cover products.

The current Landsat ARD tiles used for LCMAP’s work now serve as a starting point for NLCD 2019’s compositing work. Working with LCMAP to sift through ARD tiles for cloud-free satellite imagery with the recently-installed Denali high-performance computing system at EROS not only sped up and simplified the process, Dewitz said, but took the task of reprojecting raw Landsat data into a nationwide Albers projection – as was done prior to the advent of LCMAP - off the NLCD production team’s plate. It also cut down on the chance for errors.

 “They had already processed the data that had come right off the satellite,” Dewitz said. “We were able to go in there through Denali and take what we needed.”

The new production techniques have helped free up time for the detailed impervious surface products that are unique to the NLCD product suite. Accurately mapping a small number of new urban areas doesn’t lend itself well to automated classification, as patches of trees and bare urban areas can so easily be tagged as forest or bare ground. The human eye is still an essential element of those products, Sohl said.

“There are some things you just can’t do with an automated process at this point,” he said.

Sohl worked on the original NLCD in 1992, but he’s been focused in recent years on land cover modeling and leading the projections team for LCMAP. He took on a leadership role for NLCD again in 2020, and he’s been impressed by the production team’s speed and efficiency as it endeavors to update one of the most widely-used EROS datasets.

“Being on NLCD for the last few months has really helped me to appreciate the work they’re doing,” he said. “Being able to kick out a national scale map within a year is pretty impressive.”

Related Content

Filter Total Items: 2
Date published: August 30, 2019
Status: Active

Eyes on Earth Episode 3 - National Land Cover Database

Eyes on Earth is a podcast on remote sensing, Earth observation, land change and science, brought to you by the USGS Earth Resources Observation and Science (EROS) Center. In this episode, we talk about the National Land Cover Database.

...

Contacts: Janice Nelson
Date published: September 11, 2018
Status: Active

National Land Cover Database

EROS is central to the creation of the National Land Cover Database (NLCD), which is updated every five years and stands as the definitive land cover database for the United States.

Contacts: Collin Homer
Filter Total Items: 4
Date published: April 17, 2020

New Land Cover Maps Depict 15 Years of Alaska Change

The latest edition of the National Land Cover Database (NLCD 2016) for Alaska is now publicly available.

Date published: February 27, 2020

EROS contributes to New Land Cover Maps for U.S., Canada, Mexico

The three largest countries in North America share trade, climate and culture in a host of broad and specific ways. A new set of land cover maps for the U.S., Canada, and Mexico highlights a specific, long-standing scientific collaboration designed to benefit each nation.

Date published: May 24, 2019

New NLCD products to be folded into Esri software, Earth Engine

The National Land Cover Database (NLCD) 2016 product suite was recently released for bulk download on the Multi-Resolution Land Characteristics Consortium (MRLC) website. Users of ArcGIS and Google’s Earth Engine, however, will soon have web access to the data directly through their preferred platform.

Date published: May 20, 2019

New Land Cover Maps Depict 15 Years of Change across America

Today, the U.S. Geological Survey (USGS) released the latest edition of the National Land Cover Database (NLCD) for the U.S. – the most comprehensive land cover database that the USGS has ever produced. The NLCD 2016 documents land cover change in the Lower 48 states from 2001 to 2016. During this 15-year period, 7.6 percent of the conterminous U.S. changed land cover at least once.

Filter Total Items: 1

Mapping Roadways at Landsat Scale

A satellite flying over this area northwest of Dallas, TX will capture a wide variety of urban and rural activity. The 80-odd square miles seen in the Landsat image on the left includes airports, dense urban centers, suburban and exurban neighborhoods, manufacturing and distribution centers, oil and gas production sites, lakes, forests, wetlands and more.

Mapping

...