DHP outputs from CAN-EYE in Honghe and Hailun farms, NE China

The files in the directories include the CAN-EYE results from digital hemispherical photography (DHP) obtained over the Honghe (2012, 2013, 2019) and Hailun (2016) farms in northeastern China.

Site Description

The Honghe farm (centered at 47°39′N, 133°31′E) is located in the east of the Heilongjiang province, northeast China. This site was originally a wetland and has been converted to plant paddy rice since 1997. The paddy rice fields are flat with more than 5 km homogeneity and large rectangular fields approximately 30 m×100 m in size. A single rice variety (Japonica) is grown in this region. The rice-cropping practices are uniform, growing once a year during the summer season (May to September), with a maturation stage for about 120-150 days. The dates for the panicle formation stage, heading stage, and maturity stage are mid-June, mid-July, and early August, respectively. Paddy fields are irrigated with ground water throughout the season. The soil surface is under flooded conditions during most of the growing periods. Field measurements were performed from mid-June to mid-September 2012, from late-June to late-August 2013, and from end-June to end-August 2019.

The Hailun farm (47°24′-47°26′N, 126°47′-126°51′E) is located in the western part of the Heilongjiang province. The main crop types are maize, soybean, and sorghum. The main crops are maize, soybean, and sorghum. The maize is usually planted in early May, and the dates for tasseling, milking, and mature stages are in the end of late July, late August, and late September, respectively. The soybean and sorghum show similar growth patterns to the maize, but the soybean matures in early September. Field measurements were performed from mid-June to late September 2016.

Measurement Strategies

In Honghe,  five plots in 400 m×600 m were selected. Each plot was planted with a cultivar type and managed individually. Table 1 shows the small differences exist among the plots in terms of the plant density and plantation methods. During 2012 and 2013, in order to minimize the disruption of simultaneous destructive sampling and measurements, a moving sampling strategy was adopted. Four elementary sampling units (ESUs) of no more than 20 m×20 m in size within each plot were selected in the first week and measurements were taken for each ESU using one method. Used ESUs were discarded and another four parallel ESUs were selected for the next week measurement. In 2019, four ESUs of no more than 20 m×20 m were selected in each plot, measured once every one to two weeks, and repeated more than five times. ESU-level sampling was performed along a diamond box with two 15-meter diagonals as recommended by the VALERI network (Validation of Land European Remote Sensing Instruments, http://w3.avignon.inra.fr/valeri/). In 2012 and 2013, each ESU includes both upward view direction (the camera was placed on the ground facing the sky) and downward view direction (the camera lens was located above the canopy, facing the soil). In 2019, every ESU has only downward view direction.

Table 1. Information for the five plots in the study area.

Plot ID

Center location

Density (plants/m2)

Inter-row distance (m)

ESU (m)

A

133.515°E, 47.667°N

25

0.288

10×10

B

133.532°E, 47.663°N

26

0.286

10×10

C

133.523°E, 47.653°N

24

0.299

15×15

D

133.515°E, 47.637°N

28

0.283

15×15

E

133.534°E, 47.637°N

28

0.274

15×15

In Hailun, the study plots were flat and homogeneous and more than 100×500 m2 in size. Table 2 shows the detailed information about crop type and location in each plot. Three 20 m×20 m ESUs were set in each plot. The ESUs in the same plot were located in one row (plot A, B and C) or at the three vertices of a triangle (plot D and E). ESUs were at least 5 m away from the field borders. The three ESUs in each plot were measured every 7 days and repeated 14 times. Before July 4th, the measurement in every ESU used the downward view direction, and after that, the measurement was made using the upward view direction.

 Table 2. The detailed information about crop type and location in each plot.

Plot ID

Crop

Location

A

Maize

126.838°E, 47.410°N

B

Soybean

126.838°E, 47.405°N

C

Soybean

126.805°E, 47.401°N

D

Maize

126.798°E, 47.409°N

E

Sorghum

126.801°E, 47.429°N

The DHP images were taken using a Nikon D5100 camera with a 4.5 mm F2.8 EX DC circular fisheye convertor. An ultraviolet cap was used to prevent dust or rain from the lens. The total height of camera and the lens was about 16.5 cm. Two bubble levels were attached to the camera to keep it horizontal for both downward and upward viewing directions. System calibration for DHP camera was performed before measurement according to the CAN-EYE manual (version 6.3.3), in order to get the optical center and projection function of the lens.

CAN-EYE Data Processing

All valid photos (8~20) over one ESU were processed simultaneously by the CAN_EYE software (version 6.3.3, National Institute of Agronomical Research, Toulouse, France, https://can-eye.paca.hub.inrae.fr/) to extract the structural variables. The limit of image in viewing degrees used in this research (COI) was set to 60° by default. To get a balance between the computation time and images amount, angular resolution for zenith and azimuth directions were set to 10° and the solid angle used in computing the cover fraction was also set to 10°. A threshold process is necessary to separate the foliage from the soil background (downward view) or the sky (upward view). To minimize subjective errors, one operator performed all thresholding and classification processes.

Data Organization

For data at different locations and years, the directories are named as:

< Measurement location>< Measurement year>_DHP

For example, the results of CAN-EYE from DHP in 2012 Honghe can be found in the directory named:

“./Honghe2012_DHP”

For data at different times and ESUs, the files are named as:

< Plot ID >_< Date>__.xls

For example, the CAN-EYE output measured in the downward view direction at ESU4 of Plot A on June 11th can be found in the file named:

“PlotA_0611_Downward_ESU4.xls”

Point of Contact

Prof. Hongliang Fang

LREIS, Institute of Geographic Sciences and Natural Resources Research

Chinese Academy of Sciences (CAS)

11A Datun Road,

Beijing, 100101, China

Email: [e-mail]

Related to

Fang, Hongliang; Li, Wenjuan; Wei, Shanshan; Jiang, Chongya (2014): Seasonal variation of leaf area index (LAI) over paddy rice fields in NE China: Intercomparison of destructive sampling, LAI-2200, digital hemispherical photography (DHP), and AccuPAR methods. Agricultural and Forest Meteorology, 198-199, 126-141, https://doi.org/10.1016/j.agrformet.2014.08.005

[Field measurement in Honghe in 2012]

Fang, Hongliang; Ye, Yongchang; Liu, Weiwei; Wei, Shanshan; Ma, Li (2018): Continuous estimation of canopy leaf area index (LAI) and clumping index over broadleaf crop fields: An investigation of the PASTIS-57 instrument and smartphone applications. Agricultural and Forest Meteorology, 253-254, 48-61, https://doi.org/10.1016/j.agrformet.2018.02.003

[Field measurement in Hailun in 2016]

Fang, Hongliang; Zhang, Y; Wei, Shanshan; Li, Wenjuan; Ye, Yongchang; Sun, Tao; Liu, Weiwei (2019): Validation of global moderate resolution leaf area index (LAI) products over croplands in northeastern China. Remote Sensing of Environment, 233, 111377, https://doi.org/10.1016/j.rse.2019.111377

[Validation of moderate-resolution LAI products using field measurement in Honghai (2012 & 2013) and Hailun (2016) ]

Zhang, Y; Fang, Hongliang; Wang, Yao; Li, Sijia (2021): Variation of intra-daily instantaneous FAPAR estimated from the geostationary Himawari-8 AHI data. Agricultural and Forest Meteorology, 307, 108535, https://doi.org/10.1016/j.agrformet.2021.108535

 [Analysis of field measured FAPAR in Honghe in 2019]