Data Paper Zone II Versions EN4 Vol 2 (3) 2017
A global/continental dataset of monthly maximum fractions of photosynthetically active radiation (2001 – 2010)
: 2015 - 10 - 23
: 2016 - 06 - 01
: 2017 - 09 - 20
2786 6 0
Abstract & Keywords
Abstract: Fraction of photosynthetically active radiation (fPAR) is an effective indicator of regional vegetation and ecosystem changes, and a basis for estimating global carbon sources and sinks, as well as studying biodiversity. Based on MODIS L4 Global FPAR product MOD15A2 (1 km/8 days), we obtained the monthly maximum fPAR dataset for 2001 – 2010 through tiles mosaicking, format conversion and maximum value composite. The dataset provides monthly maximum fPAR data of the continents with 1 km spatial resolution, and can be used to assess plant productivity and study large-scale biodiversity gradients.
Keywords: fPAR; habitat; MODIS; biodiversity
Dataset Profile
Chinese title2001~2010年全球陆地光合有效辐射月最大值数据集
English titleA global/continental dataset of monthly maximum fractions of photosynthetically active radiation (2001–2010)
Corresponding authorGuo Shan (
Data authorsZhang Chunyan, Guo Shan, Guan Yanning, Cai Danlu, Wang Lei, Yao Wutao, Xiao Han
Time range2001–2010Geographical scopeGlobal/continental
Spatial resolution1 kmData volume28 GB
Data formatGeotiff
Data service system<>
Source of fundingCAS Informatization Project on "S&T Data Integration and Sharing"–"Integration and application of remote sensing data for resources and environment" (XXXH12504-1-12)
DatasetcompositionThe dataset consists of 10 zip files in total, each file corresponding to one year. A zip file comprises image documents and quality documents, totaling 240 data documents. The data files are recorded in the following format:
(1) MODIS_FPAR_1KM_MAX_MONTH_YYYYMM.tif is made up of image documents, with a data volume of 692 MB;
(2) MODIS_FPAR_QA_1KM_MAX_MONTH_YYYYMM.tif is the quality document, with a data volume of 692 MB.
1.   Introduction
The fraction of (absorbed) photosynthetically active radiation (fPAR) is expressed as a non-dimensional fraction of the incoming radiation intercepted by vegetation canopy in the band range of 400 – 700 nm. fPAR not only indicates the structure and coverage of vegetation and the exchange processes of related material and energy, but is also a key parameter for estimating the global carbon dioxide changes.1 The value of fPAR ranges from zero (on barren land) to one (for dense vegetation cover). In general, fPAR depends on the types and coverage of vegetation, and a high fPAR reflects dense green coverage. As fPAR is estimated using a number of spectral bands,2 it well reflects the dynamics of vegetation productivity.3
Brendan Mackey et al. believes that the animal behaviors of movement directly or indirectly depend on vegetation productivity, food supply and habitat conditions, and they proposed the Dynamic Habitat Template for monitoring Australian habitat conditions based on three indices: annual mean fPAR, annual minimum fPAR and coefficient of variation of fPAR. The three indices are used to assess the changes in vegetation production and the seasonality of habitat selection.4 The dynamic habitat index (DHI), which is based on remotely sensed fPAR, monitors the distribution of species and its changes using the relationships between species richness and productivity. DHI has been used in North America and Australia.3,4–8 researches indicate that spatially and seasonally induced DHI variation is closely related to the patterns of avian species richness with a correlation coefficient of up to 0.88.9 Furthermore, when coupled with long-term climate changes, DHI can be used to assess the variability of plant biodiversity.10 All this highlights the importance of monthly maximum fPAR in DHI extraction.
To date, there have been global fPAR datasets including: (1) MOD15A2 from MODIS is provided every 8 days at 1×1 km2 spatial resolution as a gridded level-4 product in the sinusoidal projection. MODIS data collection began on Day 49 of 2000. (2) fPAR datasets from GIMMS (Global Inventory Modeling and Mapping Studies), based on the imagery obtained from the Advanced Very High Resolution Radiometer (AVHRR) instrument onboard the NOAA satellite series 7, 9, 11, 14, 16 and 17, is provided every 15 days at 8×8 km2 spatial resolution, and they have a time span from August 1981 to September 2011. However, there is a lack of global datasets for monthly maximum fPAR at 1×1 km2 spatial resolution. Taking into consideration the spatial and temporal resolutions, we produced the global fPAR dataset based on MOD15A2 data. This dataset will benefit the assessment of biodiversity based on remotely sensed indexes, and promote the application of remote sensing data in ecology.
2.   Data collection and processing
2.1   Basic data
fPAR time series from MODIS (MOD15A2, 2001 – 2010), which can be attained from the NASA website (, are used to extract the monthly maximum fPAR data. The fPAR data are composited every 8 days at 1×1 km2 spatial resolution as a gridded level-4 product in the sinusoidal projection. There are 286 tiles that cover the global continents, and the track ranges are h∈(0.35) and v∈(0.17).
2.2   Data processing
The dataset underwent four steps of processing: tiles mosaicking, format conversion, data extraction, and image masking.
Step 1. Tiles mosaicking: each MODIS tile covered about 1200×1200 km2. MODIS Reprojection Tool (MRT) was used for batch processing.
Step 2 . Format conversion: after Step 1, images were in HDF format, which need be converted to the standard ENVI format using ENVI/IDL for subsequent processing.
Step 3 . Data extraction: classify the data according to month and year, and extract the monthly maximum fPAR data and the quality data in IDL©6.4 (Interactive Data Language) and ENVI©4.4 (Environment for Visualizing Images).
Step 4 . Image masking: the valid range of fPAR was from 0 to 100 (the scale factor is 0.01). For invalid values above 100, the pixels were set to 0 and the monthly fPAR data were obtained.
At first, we classified the 46 images of a year into 12 months, and extracted the monthly maximum data using IDL programs. Due to huge data storage, we didn’t compute the fPAR with the scale factor (0.01). For quality assessment, quality files were generated, for which we compared the monthly maximum data with the original data and set the pixels of quality data with the values of pixels in original data which are equal to the monthly maximum dataset. Figure 1 shows the technical flow of the data processing.

Figure 1   Technical flow for processing the global monthly maximum fPAR data
3.   Sample description
3.1   Naming format
The data files are named as follows:
(1) Image files: MODIS_FPAR_1KM_MAX_MONTH_YYYYMM.tif is an image file, with a data volume of 692 MB, where MAX_MONTH means the maximum value of a specific month of the year, and YYYYMM represents year and month.
(2) Quality files: MODIS_FPAR_QA_1KM_MAX_MONTH_YYYYMM.tif is a quality file, where QA indicates quality data.
3.2   Data sample
fPAR indicates the coverage of vegetation, and can be an indicator of vegetation productivity. In theory, dense vegetation coverage means a high fPAR, so the vegetation shows different fPAR values in different growing seasons. As showed in Figure 2, there are different fPAR patterns in different months, and white indicates oceans or regions without vegetation cover.

Figure 2   Geographical distribution of monthly maximum fPAR in 2001
4.   Quality control and assessment
The monthly maximum fPAR data are based on MOD15A2, so are the quality data. The quality data are in 8 bits integer, and each bit represents a different state as showed in Table 1. We suggest QA= 83 as the threshold, so QA < 83 indicates high quality data while the others are of low quality.11
Table 1   The quality of monthly maximum fPAR based on MOD15A2
Number of bitsParameterValueImplication
0MODLAND_QC bits0Good quality (main algorithm with or without saturation)
1Other quality (back-up algorithm or fill values)
2Detector0Detectors apparently fine for up to 50% of channels 1 & 2
150% of the detectors dead, compensated by adjacent detectors
3–4Cloud state00Significant clouds NOT present (clear)
01Significant clouds present
10Cloud present on part of the pixel
11Cloud not defined, assumed clear
(Science Computing Facilities_Quality Class flag)
000Main (RT) method used, best result possible (no saturation)
001Main (RT) method used with saturation. Good,very usable
010Main (RT) method failed due to bad geometry, empirical algorithm used
011Main (RT) method failed due to problems other than geometry, empirical algorithm used
100Pixel not produced at all, value couldn’t be retrieved (possible reasons: bad L1B data, unusable MODAGAGG data)
Reference: <>.
5.   Value and significance
The dynamic habitat index (DHI) is a composite vector deduced from monthly maximum fPAR. DHI includes Cumulative Annual Productivity (DHI-cum), Minimum Annual Apparent Cover (DHI-min), and Seasonal Variation of Greenness (DHI-sea).3 The monthly maximum fPAR values composited every 8 days at 1×1 km2 spatial resolution are used as the basic input data for calculating the three annual indices.
(1) Cumulative Annual Productivity (DHI-cum), which is estimated by summing up all monthly maximum fPAR values for each year, indirectly explains species distribution and abundance, because it characterizes the supply of resources like food supply. This is an annual total of all monthly productivity contributions:
\[DHI-cum={\sum }_{month}{MAX}_{layer,fPAR}\]
(2) Minimum Annual Apparent Cover (DHI-min) represents the lowest level (minimum value) of vegetation coverage in a year. It indicates the capacity of the landscape to sustain an adequate level of food and habitat resources during this period. This annual value represents the monthly minimum productivity of the year:
\[DHI-min=MIN\left \{ (MAX_{layer,fPAR})_{month...}\right \}\]
(3) Seasonal Variation of Greenness (DHI-sea) refers to the annual variation of natural resources associated with habitat quality, such as food, water and nutrients, which reveals the traits of life activities. This annual index represents a coefficient of variability to characterize the month-to-month standard deviation in relation to its annual mean:
\[DHI-sea=\frac{{STD\left\{{\left({MAX}_{layer,fPAR}\right)}_{month\dots }\right\}}_{}}{MEAN\left\{{{\left({MAX}_{layer,fPAR}\right)}_{month}}_{\dots }\right\}}\]
where layer is an index representing individual fPAR data set of 8-day temporal resolution, and month represents a specific month of the year. MIN, MEAN and STD are the minimum, mean and standard deviation of monthly maximum values in a particular year. Figure 3 shows the global geographical distribution of DHI, where areas with high DHI-cum and DHI-min but low DHI-sea support a relatively higher level of species richness due to rich food and habitat resources, such as the dark green areas in Figures 3a & 3b and the dark purple areas in Figure 3c (the equator), whereas areas with low DHI-cum and DHI-min but high DHI-sea lack food and habitat resources, such as the dark orange areas in Figures 3a & 3b and the dark green areas in Figure 3c (north of 60°N).

Figure 3   Geographical distribution of the dynamic habitat indices of 2001
Vegetation productivity has become a subject of increasing interest, which indicates the ability of the terrestrial ecosystem, and is an important index to assess the state of the ecosystem and the capacity of the earth.12 Species richness is proportional to vegetation productivity. Species distribution is restricted by vegetation distribution while animal behaviors are regulated by vegetation seasonality. Based on the monthly maximum fPAR dataset, more ecological indexes can be computed to support analysis on regional or national ecology.
We would like to thank NASA for providing the raw data, and Liu Xuying, Qian Dan, An Xudong and Kang Lihua for their contribution in data processing and database building.
Chen Z, Shao Q, Liu J et al. Estimating photosynthetic active radiation using MODIS atmosphere products. Journal of Remote Sensing 16 (2012): 25 – 37.
Tian Y, Zhang Y, Knyazikhin Y et al. Prototyping of MODIS LAI and fPAR algorithm with LASUR and Landsat data. IEEE Transactions on Geoscience and Remote Sensing 38 (2000): 2387 – 2401.
Coops N, Wulder M, Duro D et al. The development of a Canadian dynamic habitat index using multi-temporal satellite estimates of canopy light absorbance. Ecological Indictor 8 (2008): 754 – 766.
Mackey B, Bryan J & Randall L. Australia’s dynamic habitat template 2003. The Australian National University Annual Report2004. Available at DSpace.
Berry S, Mackey B & Brown T. Potential applications of remotely sensed vegetation greenness to habitat analysis and the conservation of dispersive fauna. Pacific Conservation Biology 13 (2007): 120 – 127.
Coops N, Fontana F, Harvey G et al. Monitoring of a national-scale indirect indicator of biodiversity using a long time-series of remotely sensed imagery. Canadian Journal of Remote Sensing 40 (2014): 179 – 191.
Coops N, Waring R, Wulder M et al. Bird diversity: a predictable function of satellite-derived estimates of seasonal variation in canopy light absorbance across the United States. Journal of Biogeography 36 (2009): 905 – 918.
Coops N, Wulder M & Iwanicka D. Demonstration of a satellite-based index to monitor habitat at continental-scales. Ecological Indictor 9 (2009): 948 – 958.
Andrew M, Wulder M, Coops N et al. Beta-diversity gradients of butterflies along productivity axes. Global Ecology and Biogeography 21 (2011): 352 – 364.
Holmes K, Nelson T, Coops N et al. Biodiversity indicators show climate change will alter vegetation in parks and protected areas. Diversity 5 (2013): 352 – 373.
Hobi M, Dubinin M, Graham C et al. A comparison of dynamic habitat indices derived from different MODIS products as predictors of avian species richness. Remote Sensing Environment 195 (2017): 142 – 152.
Yuan W, Cai W, Liu D et al. Satellite-based vegetation production models of terrestrial ecosystem: an overview. Advances in Earth Science 29 (2014): 541 – 550.
Data citation
1. Zhang C, Guo S, Guan Y et al. A global/continental dataset of monthly maximum fractions of photosynthetically active radiation (2001 – 2010). Science Data Bank. DOI: 10.11922/sciencedb.473
Article and author information
How to cite this article
Zhang C, Guo S, Guan Y et al. A global/continental dataset of monthly maximum fractions of photosynthetically active radiation (2001 – 2010). China Scientific Data 3 (2017), DOI: 10.11922/csdata. 170.2015.0025
Zhang Chunyan
data processing and analysis.
PhD, Assistant Professor; research area: remote sensing of ecological and environmental changes.
Guo Shan
data design and analysis.
Professor; research area: land-surface interaction and effects.
Guan Yanning
data design and analysis.
Professor; research area: climate change and land-surface system.
Cai Danlu
data processing and analysis.
PhD, Assistant Professor; research area: climate change and surface vegetation.
Wang Lei
data download and preprocessing.
Master's student; research area: remote sensing of urban environment.
Yao Wutao
data download and preprocessing.
Master's student; research area: remote sensing of urban environment.
Xiao Han
data download and preprocessing.
Master's student; research area: remote sensing of ecological and climate changes.
CAS Informatization Project on "S&T Data Integration and Sharing" – "Integration and application of remote sensing data for resources and environment" (XXXH12504-1-12)
Publication records
Published: Sept. 20, 2017 ( VersionsEN4
Released: June 1, 2016 ( VersionsZH1
Published: Sept. 21, 2017 ( VersionsZH2