GET THE APP

Method for Burned Forest Biomass Estimation at Subcompartment Lev
Journal of Geography  & Natural Disasters

Journal of Geography  & Natural Disasters
Open Access

ISSN: 2167-0587

+44-20-4587-4809

Research Article - (2016) Volume 6, Issue 3

Method for Burned Forest Biomass Estimation at Subcompartment Level Using GF-1 images and GIS Datasets

Xianlin Qin*, Lingyu Ying, Guifen Sun and Xiaofeng Zu
Research Institute of Forest Resource Information Technique, Chinese Academy of Forestry, Beijing, 100091, PR China
*Corresponding Author: Xianlin Qin, Research Institute of Forest Resource Information Technique, Chinese Academy Of Forestry, Beijing, 100091, PR China, Tel: +86-010-62888847, Fax: +86-010-62888315 Email:

Abstract

To estimate the burned biomass and get the burned fuel types information by forest fire for the need of Chinese forestry management, basing on the fuel load from digital reference data, the combustion factor gotten from fieldwork, and the results of burned scar mapping by using the No.1 High-Resolution satellite (GF-1) of Chinese, the burned biomass estimation method at the subcompartment level has been developed using satellite images and Geography datasets. The method has been validated by the selected forest fire, which had taken place in Huangcaobai of Anning City, Yunnan province in year 2012.The total burned biomass is about 1.18 × 108 kg by using the panchromatic and multispectral scanners (PMS) image of GF-1; however, it is about 1.11 × 108 kg by using the Wide Coverage Image (WFV) of GF-1. The difference between them is 7.10 × 106 kg. This study also supplies a method for the single forest fire case when the fire radiative power (FRP) or fire radiative energy (FRE) of detected active fire points by using sparse low spatial resolution satellite images doesn’t satisfied the condition of Power Law distribution or Gaissian function.

Keywords: Burned forest biomass; GF-1; Forest fire; Remote sensing technique

Introduction

Vegetation biomass burning has been identified as a significant source of aerosols, carbon fluxes, and trace gases, which pollute the atmosphere and contribute to radiative forcing responsible for global climate change [1,2]. The trace gas from vegetation biomass burning has been paid more attention from the scientific community over the past several decades as an important contributor to total climatic radiative forcing [3]. Spatial and temporal explicit mapping of the amount of burned biomass by fire is needed to estimate atmospheric emissions of greenhouse gases and aerosols that have a significant climate forcing effect. The methods for burned biomass estimation have been recently developed according to the assumption that the fire radiative power (FRP) or fire radiative energy (FRE) satisfies the Power Law distribution function or Gassian function [4-9]. In fact, the active fire points are so spare by using the low spatial resolution polar orbit satellite images [10-13]. The spare FRP or FRE of a single forest fire case usually doesn’t satisfy the Power Law distribution function or Gassian function. In this case, it isn’t suitable to estimate the burned forest biomass by using the methodology based on the model of Power Law or Gassian. In China, the basic forestry management is at subcompartment level. The information of the burned fuel type, burned biomass, and damaged is need for the forest manager at the subcompartment level after the forest fire. To get the method for burned biomass estimation at subcompartment level, the Chinese satellite images and geography datasets have been selected as the data for the study.

Study area and Materials

Study area

The burned biomass of the same forest fire, which had taken place in Huangcaobai of Anning City, Yunnan province in March 19, 2012, has been selected as the experimental area for the method implement. The location of this study is shown in Figure 1.

geography-natural-disasters-Experimental-area

Figure 1: Experimental area.

The forest fire had lasted about 5 days. The local forest includes Dipan pine (Pinus yunnanensis var. pygmaea), Yunnan pine (Pinus yunnanensis), and Fir (Keteleeria fortunei (Murr.) Carr).

Materials

The No.1 high-resolution satellite (GF-1) is the first series of Chinese high-resolution satellites, which had been successfully launched on April 23, 2013. Its’ panchromatic and multispectral scanners (PMS) camera can get the 8 m multispectral image with blue, green, red and near-infrared band and 2 m panchromatic images. It also equipped with the Wide Coverage Image (WFV), which spatial resolution is 16 m. The main parameters of the GF-1 PMS and GF-1 WFV have been listed in Table 1.

Parameters PMS cameras WFV cameras
Wavelength Panchromatic 0.45-0.90μm /
Multi-spectral Scaner 0.45-0.52μm (Band 1) 0.45-0.52μm (Band 1)
0.52-0.59μm (Band 2) 0.52-0.59μm (Band 2)
0.63-0.69μm (Band 3) 0.63-0.69μm (Band 3)
0.77-0.89μm (Band 4) 0.77-0.89μm (Band 4)
Spatial resolution Panchromatic 2m /
Multi-spectral Scaner 8m 16m
Frame width 60km (combing 2 cameras) 800km (combing 4 cameras)
Recycle (Shifting) 4 days /
Recycle(No shifting) 41days 4 days

Table 1: The parameters of GF-1 satellite image.

To compare the difference by using different spatial resolution satellite images, the GF-1 PMS and GF-1 WFV images which cover the forest fire have been selected to the burned area mapping. The GF-1 WFV image of September 29, 2013 and the GF-1 PMS image of February 14, 2014 has been selected as the satellite image to get the burned area edge respectively because the burned area doesn’t be covered by cloud. At the same times, to get the burned fuel type and biomass, the digital maps of fuel type and fuel load at subcompartment level have been collected from local institution.

Methods

After the reprocessing to the selected GF-1 WFV and GF-1 PMS data, the burned area has been respectively extracted using GF-1 WFV and GF-1 PMS data. The reprocessing includes radiometric calibration, orthorectification, atmospheric correction and resolution merger. Orthorectification using Rational Polynomial Coefficient (RPC) and Digital Elevation Model (DEM); Atmospheric correction using Fast Line-of-sight Atmospheric Analysis of Hypercubes (FLAASH) software; and resolution merger using Gram-Schmidt Pan Sharpening method for the Panchromatic and Multi-spectral Scaner of GF-1 PMS sensor in ENVI 5.1 software. The burned biomass can be calculated when the burned area, the fuel load and the combustion factor of every subcompartment have been known. The burned area can be gotten by using the selected GF-1 images. The digital fuel type and fuel load at subcompartment level have been collected from local institution. The flowchart of this study is shown in Figure 2.

geography-natural-disasters-Flowchart-illustrates

Figure 2: Flowchart illustrates the methodology of the study.

Burned area extraction

The burned scar edge has been respectively extracted by using the GF-1 PMS image and GF-1 WFV in this study. The selected method for the burned scar edge extraction has been developed by Xiaofeng Zu, et al. [14]. Then, the burned area at subcompartment level can be gotten through the spatial analysis with the burned edge coming from satellite and the edge of subcompartment in ArcGIS 9.3 software.

Combustion factor estimation

The combustion factor of every subcompartment has been measured through filed work in this study. Their range has been assigned from 0 to 1.0, with an interval 0.1 according to the burned area and the severity of the subcompartment. The value is 0 for unburned subcompartment and 1.0 for complete burned.

Model

The model for burned biomass estimation at subcompartment level has been developed as Equation 1 based on the model of Seiler’s [15].

Mi = Ai × Bi × Fi (1)

Where: Mi is the burned biomass (kg) of the subcompartment i; Ai is the burned area (m2) of subcompartment i; Bi is the fuel load (kg/m2) of subcompartment i; and Fi is combustion factor (fraction of available burned fuel) of subcompartment i.

Then, the total burned biomass can be calculated using Equation 2.

Equation (2)

Where: M is the total burned biomass (kg); n is the number of sub-compartment.

Results

The results of burned area mapping by using the GF-1 PMS and GF-1 WFV have been showed in Figure 3. The results of burned biomass according to the method of section 5.3 and using the same classification value have been showed in Figure 4. It clearly shows the distribution of burned biomass from Figure 4, which can help the forest managers know the information of damaged forest and arrange the management in future.

geography-natural-disasters-burned-area

Figure 3: The results of burned area mapping (a) Using GF-1 WFV image; (b) Using GF-1 PMS image.

geography-natural-disasters-Burned-biomass

Figure 4: Burned biomass distribution map (a) using GF-1 PMS image; (b) using GF-1 WFV image.

Discussion

The effection of burned area mapping

The results of burned area mapping had been compared by using the GF-1 PMS image and GF-1 WFV image. To get the different between the burned area and the active fire points, the active fire points, which had been gotten from the Terra/Aqua MODIS active fire production (MOD14A1/ MYD14A1) during the forest fire, have been analyzed with the burned area mapping in ArcGIS 9.3 software. The results showed in Figure 5. As we can see from the Figure 5, the extracted burned scar edge was clearly different by using the GF-1 PMS and the GF-1 WFV image. The spatial resolution of GF-1 PMS is 2/8 meter at nadir; However, it is 16 meter for the GF-1 WFV image. The validation showed that the accuracy of extracted burned scar edge by using the GF-1 PMS image is higher than that of GF-1 WFV image. There is a large burned area that didn’t covered by the active fire points from Figure 5.

geography-natural-disasters-comparing-burned

Figure 5: The result for comparing the burned area and active fire points.

At the same times, we also analysed the frequence of FRP of the active fire points. There were only 24 active fire points by using Terra/ Aqua MODIS during the five days. The results showed in Figure 6. It shows that the distribution of FRP of the selected forest fire doesn’t satisfy the condition of Power Law distribution or Gassian. So, the burned biomass estimated methodology based on the distribution of Power Law or Gassian function couldn’t be properly applied in the forest fire case.

geography-natural-disasters-Aqua-MODIS

Figure 6: The frequency of FRP of active fire points using Terra/Aqua MODIS image.

Burned biomass analysis

The main burned fuel type and the difference of burned biomass had also been analysed. The relative error about the estimated burned biomass between by using the GF-1 PMS image and GF-1 WFV image has been also calculated using Equation 3. Because there are’t the truth value of the burned biomass, the estimated burned biomass by using GF-1 PMS image has been selected as the based burned biomass. The R has been setted as nodata when A is 0 when. The results showed in Table 2.

Fuel Type GF-1 PMS(kg) GF-1 WFV(kg) Difference(kg) R(%)
Yunnan Pine 0.00 11008.00 11008.00 nodata
Dipan Pine 94608843.17 90541690.42 -4067152.75 -4.30
Economic Forest 553096.81 2655.35 -550441.46 -99.52
Shrub 18939224.23 15843165.94 -3096058.29 -16.35
Grass 1742097.00 1664911.52 -77185.48 -4.43
Others 2513504.42 3189824.14 676319.72 26.91
Total 118356765.63 111253255.37 -7103510.26 -6.00

Table 2: The results of burned biomass estimation according to fuel types.

Equation (3)

Where: R is the relative error; A is the estimated burned biomass by using GF-1 PMS image; D is the difference of burned biomass of fuel type or subcompartment by using the GF-1 PMS image and GF-1 WFV image.

It shows that the main burned biomass is Dipan Pine from Table 2. Both of them are more than 9,000,000 kg.There are large difference between using GF-1 PMS and GF-1 WFV images from Table 2. The total difference between the value using the GF-1 PMS image and GF-1 WFV iamge is 7.10 × 106 kg. In this study, the fuel load coming from the reference data and the combustion factor gotten by filed work at subcompartment level; the burned scar edge has been extracted by using GF-1 PMS and GF-1 WFV images respectively. So, the difference comes from the different result of burned scar mapping.

It shows that the highest on R is Economic Forest from Table 2, with -99.52%. The second is others, with 17.46%. In addition, the lowest on R is Dipan Pine, with 4.58%. The R of the total difference is -6.00%. At the same times, the burned biomass has been also analyzed according to the subcompartment. Part of the results showed in Table 3 (because there are 249 subcompartment for the forest fire).

Subcompartment Number GF-1 PMS(kg) GF-1 WFV(kg) Difference(kg) R(%)
4156 251157.76 256055.03 4897.27 1.95
4300 0.00 4041.92 4041.92 /
4307 107000.44 140208.23 33207.79 31.04
4313 305305.61 310595.12 5289.51 1.73
4350 123164.61 120621.01 -2543.60 -2.07
4353 55500.69 165249.76 109749.07 197.74
4356 144927.56 323715.94 178788.38 123.36
4360 859584.17 915335.96 55751.80 6.49
4377 49496.14 34499.33 -14996.80 -30.30
4382 45586.46 102786.60 57200.14 125.48
4482 5184.05 24391.01 19206.95 370.50
4854 652313.39 652889.81 576.42 0.09
Total 118356765.63 111253255.37 -7103510.26 -6.00

Table 3: Part of burned biomass estimation results according to subcompartment.

The estimated burned biomass according to the subcompartment is obvious difference between by using the GF-1 PMS and GF-1 WFV image from Table 3. The highest on R is the No. 4482 subcompartment, with 370.50%. The lowest on R is the No. 4854, with the 0.09%.

Conclusions

The burned forest biomass has been estimated by using the burned area, the fuel load and the combustion factor based on the subcompartment level. It shows that the method can supply the forestry manager the information of burned fuel type, burned biomass and burned area of every subcompartment by the forest fire for forestry management. At the sometimes, it supplies a method for the single forest fire case when the FRP or FRE of active fire points, getting from the sparse low spatial resolution satellite images, doesn’t satisfied the condition of Power Law distribution model or Gassian function.

However, there are some disadvantages of the methodology because the uncertainty about the burned scars edge. The accuracy of burned area mapping is a key factor to the burned biomass estimation. It is the better way to select the high spatial resolution satellite images to map the burned area than using the low spatial resolution satellite images.

Acknowledgments

This work was supported by the Chinese space program pre-research project, the national science and technology major projects of China (21-Y30B05-9001- 13/15), and the Dragon 3 project (10350).

Author Contributions

Xianlin Qin conceived and designed the experiments, field work, developed the burned biomass estimation method, and wrote the paper. Xiaofeng Zu mapped the burned area using GF-1 images. Lingyu Ying analyzed the FRP distribution. Guifen Sun has taken part in the validation for the burned biomass.

Conflicts of Interest

All of us declared no conflict of interest in our study and in this paper.

References

  1. Crutzen PJ, Andreae MO (1990) Biomass burning in the tropics: Impact on atmospheric chemistry and biogeochemical cycles. Science 250: 1669-1678.
  2. Liane SG, Boone KJ, Warren BC (2004) Modeling biomass burning emissions for Amazon forest and pastures in Rondia, Brazil. Ecological Applications, Supplement 14: 232-246.
  3. Freeborn PH, Wooster MJ, Hao WM (2008) Relationships between energy release, fuel mass loss, and trace gas and aerosol emissions during laboratory biomass fires. J Geophys Res 113: 1-17.
  4. Andreae MO, Merlet P (2001) Emission of trace gases and aerosols from biomass burning. Global biogeochemical cycles 15: 955-966.
  5. Wooster MJ, Roberts G, Perry GLW, Kaufman YJ (2005) Retrieval of biomass combustion rates and totals from fire radiative power observations: FRP derivation and calibration relationships between biomass consumption and fire radiative energy release. J Geophys Res 110.
  6. Roberts GJ, Wooster MJ, Perry GLW (2005) Retrieval of biomass combustion rates and totals from fire radiative power observations: Application to southern Africa using geostationary SEVIRI imagery. J Geophys Res 110: D21111.
  7. Ellicott E, Vermote E, Giglio L, Roberts G (2009) Estimating biomass consumed from fire using MODIS FRE. Geophys Res Lett 36: L13401.
  8. Vermote E, Ellicott E, Dubovik O (2009) An approach to estimate global biomass burning emissions of organic and black carbon from MODIS fire radiative power. J Geophys Res 114: D18205.
  9. Kumar SS, Roy DP, Boschetti L, Kremens R (2011) Exploiting the power law distribution properties of satellite fire radiative power retrievals: A method to estimate fire radiative energy and biomass burned from sparse satellite observations. J Geophys Res 116: D19303.
  10. Giglio L, Descloitres J, Justice CO, Kaufman YJ (2003) An enhanced contextual fire detection algorithm for MODIS. Remote Sens Environ 87: 273-282.
  11. Qin XL, Zhang ZH, Li ZY (2008) Forest fire monitoring using ENVISAT-AATSR and MEIRS Images. Proc. of the '2nd MERIS/(A)ATSR User Workshop', Frascati, Italy, 22-26 September (ESA SP-666).
  12. Qin XL, Zhu X, Yang F (2013) Analysis of sensitive spectral bands for burning status detection using Hyper-Spectral images of Tiangong-01. Spectroscopy and Spectral Analysis 33: 1908-1911.
  13. Qin XL, Yang F, Zu XF (2014) Quantitative extraction of fine contour parameters for forest fire using satellite remote sensing. J Infrared Millim Waves 33: 642-648.
  14. Zu XF, Qin XL, Yin LY (2015) Study on the burned area based on the decision tree classification techniques with vegetation index. Forest resources management 4: 73-78.
  15. Seiler W, Crutzen PJ (1980) Estimates of gross and net fluxes of carbon between the biosphere and the atmosphere from biomass burning. Climatic change 2: 207-247.
Citation: Qin X, Ying L, Sun G, Zu X (2016) Method for Burned Forest Biomass Estimation at Subcompartment Level Using GF-1 images and GIS Datasets. J Geogr Nat Disast 6:181.

Copyright: © 2016 Qin X, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Top