USDA LTAR Common Experiment measurement: PhenoCam for Green Chromatic Coordinate
Dawn M. Browning, Nicanor Z. Saliendra, Craig W. Whippo
phenology
image analysis
camera
photograph
vegetation
canopy
normalized difference vegetation index
green chromatic coordinate
Long-Term Agroecosystem Research
LTAR
crops
grazingland
rangeland
Common Experiment
Disclaimer
This research is a contribution from the Long-Term Agroecosystem Research (LTAR) network. LTAR is supported by the United States Department of Agriculture. The use of trade, firm, or corporation names in this publication (or page) is for the information and convenience of the reader. Such use does not constitute an official endorsement or approval by the United States Department of Agriculture or the Agricultural Research Service of any product or service to the exclusion of others that may be suitable. USDA is an equal opportunity provider and employer.
Abstract
Phenology is the study of seasonal dynamics and how environmental conditions such as temperature and rainfall influence it over different time periods. In the case of plants, phenology is a primary control of productivity. Phenology can be studied through color-based image analysis using a digital camera (PhenoCam) most commonly mounted on a tower that captures time-lapse photographs of vegetation canopies. Metrics of canopy greenness can be calculated from PhenoCam imagery as a continuous record of canopy and surface greenness. The most commonly used greenness metric is the green chromatic coordinate (GCC), which is similar to the normalized difference vegetation index derived from satellite remote sensing. PhenoCam images provide unique insights into vegetation dynamics at a daily time step and the associated ecosystem metrics affected by management and climate change.
Steps
Protocol
Camera installation, operation, and maintenance
A Stardot camera (model NetCam-SC-IR) is installed at meteorological and flux tower locations at an appropriate height above the vegetation canopy, facing north to minimize sun angle effects throughout the year.
When selecting the location to install a camera (if not on an established tower), the best practice is to ensure the landscape the camera is focused on (i.e., the field of view) represents the surrounding landscape across which measurements are made.
Landscape views are most common and work well with flux tower applications. Some (relatively few) smaller-scale, experimental PhenoCam installations focus on small plots or individual plants. These installations are prone to challenges with shadows and limit extrapolation of results.
For a given installation, the camera is oriented approximately 15° off the horizon so that more foreground than sky is in the field of view, i.e., a small portion of the sky is viewed at the top of an image.
The camera is connected to a laptop PC, and the PhenoCam installation tool (PIT) is run to automatically capture red–green–blue (RGB) and IR images every 30 minutes.
Images and metadata files are transferred via secure protocol to the PhenoCam Network at Northern Arizona University.
Clean the glass on the housing with a lint-free cloth and distilled or deionized water. Do not wipe the glass surface when dry. After removing dirt from the glass with water, clean the glass with 95% ethanol and clean lint-free cloth. Complete the cleaning process by removing the remaining dust or lint with a can of compressed air. Repeat this cleaning procedure quarterly or as needed, particularly after field operations such as planting and harvesting.
Image metadata
Image metadata include camera name, camera type, datetime, and camera temperature. The PhenoCam Network populates metadata. The investigator provides additional metadata, such as primary species and latitude and longitude, when the site is established.
Calculations
The camera captures the RGB additive color seen by the human eye. The color and brightness of a given pixel are recorded as a digital number (DN).
The green chromatic coordinate (GCC) is calculated as the quotient of the intensity of green relative to the total intensity of red, green, and blue:
GCC = GDN / (RDN + GDN + BDN) [1]
where GDN, RDN, and BDN are the digital numbers (DN) of green, red, and blue, respectively summarized for DN values within the region of interest (ROI).
The PhenoCam Network or site scientist can annotate the ROIs.
An ROI’s GCC is calculated as part of the PhenoCam Network data production protocol.
Quality assessment and quality control
Images are evaluated by QA/QC test per PhenoCam Network protocols.
Protocols established by the PhenoCam Network automatically notify site contacts after three, five, and ten days without receiving image data. Site scientists or field personnel should evaluate and remedy field conditions in this case. Site personnel are encouraged to check the status of the image time series on the PhenoCam Network regularly (daily or weekly) for streaming and image quality. The latest images are shown here: https://phenocam.nau.edu/webcam/sites/sitename/, where “sitename” is the PhenoCam site name. Site cameras can be found by searching the site name here: https://phenocam.nau.edu/webcam/network/search/.
Image archiving
The PhenoCam Network currently sustains the Long-Term Agricultural Research (LTAR) network PhenoCam image archive, metadata, and data products using two strategies: redundancy and backup. The primary archive is housed in a climate-controlled and access-restricted machine room, and a mirrored backup is maintained on NAU’s Monsoon cluster.
In addition, LTAR has established a mirror of image data and derived data products from the PhenoCam Network to the USDA SCINet Juno servers as a secure backup and archive, backed up every two days.
Illustrative Information
Primary productivity (photosynthesis) and GCC are strongly correlated among ecosystems, including forests, shrublands, and grasslands (Fig. 6; Richardson, 2023b). In conjunction with eddy covariance measurements, GCC can be used to predict water and carbon fluxes as influenced by contrasting management practices within and across agroecosystems (Browning et al., 2021).
#尊敬的用户，由于网络监管政策的限制，部分内容暂时无法在本网站直接浏览。我们已经为您准备了相关原始数据和链接，感谢您的理解与支持。
https://lh7-us.googleusercontent.com/BqyDqhjGlVKAuavFm9wjQ4dVFuVHH3sWv-PyculJ9I0d85Rse9iFsFeeb9raXV8aKgaLuMJwwEnfTsatEyMOow6alpHT0YUGImze82VCjwG1w3vdBG30DD7D5KH8N4UZfrdFrVHaV69wSTcn9by_AQ
Accessing PhenoCam Data
Version 2 of the PhenoCam dataset has been published (Seyednasrollah et al., 2019).* Richardson (2023a) describes a very detailed tutorial tutorial on how to access PhenoCam data.
- Daily GCC and other products are accessed through a ”Gallery” of PhenoCam sites; a site is searched and found by geography (map) or name.
- Within a site’s webpage is a link to the “ROI Timeseries” that is selected (mouse-clicked) to view a graph of daily GCC, as exemplified in Fig. 7.
- Selecting/clicking the link to “Provisional Data” opens a dialog to download a compressed file (provisional_data.zip) containing data folders and documentation (README.txt and fair use policy).
- The National Ecological Observatory Network (NEON) also provides access to PhenoCam provisional data.
#尊敬的用户，由于网络监管政策的限制，部分内容暂时无法在本网站直接浏览。我们已经为您准备了相关原始数据和链接，感谢您的理解与支持。
https://lh7-us.googleusercontent.com/26HjLzuKZclnqmsfnQzsu13hYfuxZkfriXlen5btx7oj-oXtxR6VCUR9Azl7QGsOXF8tLfdfNjfGjeESCGL20kDaJxIfdnYv3-JPipkntuOTiWYhgc71U9Bhrswk7XqktASvk5TqjqiNdd7OJYqGEQ
Recommendations for data collection
Table 1. Summary of recommendations for measurement of PhenoCam attributes.
| A | B | C | D |
|---|---|---|---|
| Attribute | Preferred | Minimum | Comments |
| Spatial scale | Plot or field; field of view or region of interest (ROI); the camera height determines the size of the ROI; a higher camera height implies a greater ROI or field of view | The camera height must be at least 2 m above the vegetation canopy | |
| Frequency | Photograph taken every 30 minutes in daytime and uploaded to the PhenoCam Network | Photograph taken every 30 minutes in daytime and uploaded to the PhenoCam Network via cellular service and router (PIT) | Other methods for transmitting images using a data logger are available here. |
| Covariate metrics | Soil moisture; eddy covariance sensors to measure fluxes of CO2 and H2O or latent heat, sensible heat; above-ground biomass, leaf area index, and canopy height | Meteorological station within 1 km to include precipitation, air temperature, and relative humidity | Meteorological station within 3 km at a minimum |
