Terrestrial primary production for the conterminous United States derived from Landsat 30 m and MODIS 250 m

Nathaniel P. Robinson; Brady W. Allred; William K. Smith; Matthew O. Jones; Alvaro Moreno; Tyler A. Erickson; David E. Naugle; Steven W. Running

doi:10.1002/rse2.74

Terrestrial primary production for the conterminous United States derived from Landsat 30 m and MODIS 250 m

Nathaniel P. Robinson
, Brady W. Allred
, William K. Smith
, Matthew O. Jones
, Alvaro Moreno
, Tyler A. Erickson
, David E. Naugle
, Steven W. Running

Research output: Contribution to journal › Article › peer-review

157 Scopus citations

Abstract

Terrestrial primary production is a fundamental ecological process and a crucial component in understanding the flow of energy through trophic levels. The global MODIS gross primary production (GPP) and net primary production (NPP) products (MOD17) are widely used for monitoring GPP and NPP at coarse resolutions across broad spatial extents. The coarse input datasets and global biome-level parameters, however, are well-known limitations to the applicability of the MOD17 product at finer scales. We addressed these limitations and created two improved products for the conterminous United States (CONUS) that capture the spatiotemporal variability in terrestrial production. The MOD17 algorithm was utilized with medium resolution land cover classifications and improved meteorological data specific to CONUS in order to produce: (a) Landsat derived 16-day GPP and annual NPP at 30 m resolution from 1986 to 2016 (GPP_L ₃₀ and NPP_L ₃₀, respectively); and (b) MODIS derived 8-day GPP and annual NPP at 250 m resolution from 2001 to 2016 (GPP_M ₂₅₀ and NPP_M ₂₅₀ respectively). Biome-specific input parameters were optimized based on eddy covariance flux tower-derived GPP data from the FLUXNET2015 database. We evaluated GPP_L ₃₀ and GPP_M ₂₅₀ products against the standard MODIS GPP product utilizing a select subset of representative flux tower sites, and found improvement across all land cover classes except croplands. We also found consistent interannual variability and trends across NPP_L ₃₀, NPP_M ₂₅₀, and the standard MODIS NPP product. We highlight the application potential of the production products, demonstrating their improved capacity for monitoring terrestrial production at higher levels of spatial detail across broad spatiotemporal scales.

Original language	English
Pages (from-to)	264-280
Number of pages	17
Journal	Remote Sensing in Ecology and Conservation
Volume	4
Issue number	3
DOIs	https://doi.org/10.1002/rse2.74
State	Published - Sep 2018

Funding

We thank the Google Earth Engine developers for their support and technical advice. This work was funded through a Google Earth Engine Research Award and by the NRCS Wildlife Conservation Effects Assessment Project and Sage Grouse Initiative. This work used eddy covariance data acquired and shared by the FLUXNET community, including these networks: AmeriFlux, Afri-Flux, AsiaFlux, CarboAfrica, CarboEuropeIP, CarboItaly, CarboMont, ChinaFlux, Fluxnet-Canada, GreenGrass, ICOS, KoFlux, LBA, NECC, OzFlux-TERN, TCOS-Siberia, and USCCC. The ERA-Interim reanalysis data are provided by ECMWF and processed by LSCE. The FLUXNET eddy covariance data processing and harmonization was carried out by the European Fluxes Database Cluster, AmeriFlux Management Project, and Fluxdata project of FLUXNET, with the support of CDIAC and ICOS Ecosystem Thematic Center, and the OzFlux, ChinaFlux and AsiaFlux offices.

Funders
Sage Grouse Initiative

Keywords

Google earth engine
MOD17
MODIS
gross primary production
landsat
net primary production

Access to Document

10.1002/rse2.74

Cite this

@article{7f51f6c273244f4d891f7efd21980724,

title = "Terrestrial primary production for the conterminous United States derived from Landsat 30 m and MODIS 250 m",

abstract = "Terrestrial primary production is a fundamental ecological process and a crucial component in understanding the flow of energy through trophic levels. The global MODIS gross primary production (GPP) and net primary production (NPP) products (MOD17) are widely used for monitoring GPP and NPP at coarse resolutions across broad spatial extents. The coarse input datasets and global biome-level parameters, however, are well-known limitations to the applicability of the MOD17 product at finer scales. We addressed these limitations and created two improved products for the conterminous United States (CONUS) that capture the spatiotemporal variability in terrestrial production. The MOD17 algorithm was utilized with medium resolution land cover classifications and improved meteorological data specific to CONUS in order to produce: (a) Landsat derived 16-day GPP and annual NPP at 30 m resolution from 1986 to 2016 (GPPL 30 and NPPL 30, respectively); and (b) MODIS derived 8-day GPP and annual NPP at 250 m resolution from 2001 to 2016 (GPPM 250 and NPPM 250 respectively). Biome-specific input parameters were optimized based on eddy covariance flux tower-derived GPP data from the FLUXNET2015 database. We evaluated GPPL 30 and GPPM 250 products against the standard MODIS GPP product utilizing a select subset of representative flux tower sites, and found improvement across all land cover classes except croplands. We also found consistent interannual variability and trends across NPPL 30, NPPM 250, and the standard MODIS NPP product. We highlight the application potential of the production products, demonstrating their improved capacity for monitoring terrestrial production at higher levels of spatial detail across broad spatiotemporal scales.",

keywords = "Google earth engine, MOD17, MODIS, gross primary production, landsat, net primary production",

author = "Robinson, \{Nathaniel P.\} and Allred, \{Brady W.\} and Smith, \{William K.\} and Jones, \{Matthew O.\} and Alvaro Moreno and Erickson, \{Tyler A.\} and Naugle, \{David E.\} and Running, \{Steven W.\}",

note = "Publisher Copyright: {\textcopyright} 2018 The Authors. Remote Sensing in Ecology and Conservation published by John Wiley \& Sons Ltd on behalf of Zoological Society of London",

year = "2018",

month = sep,

doi = "10.1002/rse2.74",

language = "English",

volume = "4",

pages = "264--280",

journal = "Remote Sensing in Ecology and Conservation",

issn = "2056-3485",

publisher = "Wiley-Blackwell Publishing Ltd",

number = "3",

}

TY - JOUR

T1 - Terrestrial primary production for the conterminous United States derived from Landsat 30 m and MODIS 250 m

AU - Robinson, Nathaniel P.

AU - Allred, Brady W.

AU - Smith, William K.

AU - Jones, Matthew O.

AU - Moreno, Alvaro

AU - Erickson, Tyler A.

AU - Naugle, David E.

AU - Running, Steven W.

PY - 2018/9

Y1 - 2018/9

N2 - Terrestrial primary production is a fundamental ecological process and a crucial component in understanding the flow of energy through trophic levels. The global MODIS gross primary production (GPP) and net primary production (NPP) products (MOD17) are widely used for monitoring GPP and NPP at coarse resolutions across broad spatial extents. The coarse input datasets and global biome-level parameters, however, are well-known limitations to the applicability of the MOD17 product at finer scales. We addressed these limitations and created two improved products for the conterminous United States (CONUS) that capture the spatiotemporal variability in terrestrial production. The MOD17 algorithm was utilized with medium resolution land cover classifications and improved meteorological data specific to CONUS in order to produce: (a) Landsat derived 16-day GPP and annual NPP at 30 m resolution from 1986 to 2016 (GPPL 30 and NPPL 30, respectively); and (b) MODIS derived 8-day GPP and annual NPP at 250 m resolution from 2001 to 2016 (GPPM 250 and NPPM 250 respectively). Biome-specific input parameters were optimized based on eddy covariance flux tower-derived GPP data from the FLUXNET2015 database. We evaluated GPPL 30 and GPPM 250 products against the standard MODIS GPP product utilizing a select subset of representative flux tower sites, and found improvement across all land cover classes except croplands. We also found consistent interannual variability and trends across NPPL 30, NPPM 250, and the standard MODIS NPP product. We highlight the application potential of the production products, demonstrating their improved capacity for monitoring terrestrial production at higher levels of spatial detail across broad spatiotemporal scales.

AB - Terrestrial primary production is a fundamental ecological process and a crucial component in understanding the flow of energy through trophic levels. The global MODIS gross primary production (GPP) and net primary production (NPP) products (MOD17) are widely used for monitoring GPP and NPP at coarse resolutions across broad spatial extents. The coarse input datasets and global biome-level parameters, however, are well-known limitations to the applicability of the MOD17 product at finer scales. We addressed these limitations and created two improved products for the conterminous United States (CONUS) that capture the spatiotemporal variability in terrestrial production. The MOD17 algorithm was utilized with medium resolution land cover classifications and improved meteorological data specific to CONUS in order to produce: (a) Landsat derived 16-day GPP and annual NPP at 30 m resolution from 1986 to 2016 (GPPL 30 and NPPL 30, respectively); and (b) MODIS derived 8-day GPP and annual NPP at 250 m resolution from 2001 to 2016 (GPPM 250 and NPPM 250 respectively). Biome-specific input parameters were optimized based on eddy covariance flux tower-derived GPP data from the FLUXNET2015 database. We evaluated GPPL 30 and GPPM 250 products against the standard MODIS GPP product utilizing a select subset of representative flux tower sites, and found improvement across all land cover classes except croplands. We also found consistent interannual variability and trends across NPPL 30, NPPM 250, and the standard MODIS NPP product. We highlight the application potential of the production products, demonstrating their improved capacity for monitoring terrestrial production at higher levels of spatial detail across broad spatiotemporal scales.

KW - Google earth engine

KW - MOD17

KW - MODIS

KW - gross primary production

KW - landsat

KW - net primary production

UR - https://www.scopus.com/pages/publications/85053592835

U2 - 10.1002/rse2.74

DO - 10.1002/rse2.74

M3 - Article

AN - SCOPUS:85053592835

SN - 2056-3485

VL - 4

SP - 264

EP - 280

JO - Remote Sensing in Ecology and Conservation

JF - Remote Sensing in Ecology and Conservation

IS - 3

ER -

Terrestrial primary production for the conterminous United States derived from Landsat 30 m and MODIS 250 m

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this