High-accuracy 3D information is quite useful in various areas. Currently, following applications are considered.
1. Observation of tree heights to estimate the biomass quantity.
2. Making the global elevation map with high resolution.
3. Observation of ice-sheets.
This paper aims at reporting the present state of our conceptual study of the GLS. A prospective performance of the GLS for earth observation missions mentioned above.
High-accuracy three-dimensional (3D) information of global area is very useful in various fields, such as global observations of tree height, elevation, and ice sheet. Especially, there are pressing needs to advance understanding of how changes in the 3D structure of terrestrial vegetation are affecting the global carbon dynamics and their implications for climate change. Thus new space assets are urgently needed to measure global maps of the 3D structure of vegetation .
The Japan Aerospace Exploration Agency (JAXA) has started a conceptual study of earth observation missions with a space-borne laser scanner (GLS, as Global Laser Scanner) which will enable us to obtain the high-accuracy 3D information from the globe. Fig. 1 shows a schematic diagram of GLS.
Laser scanner is a system which transmit intense pulsed laser light to the ground from an airplane or spacecraft, receive the scattered light, and measure the distance to the surface from the round-trip delay time of the pulse. As shown in Fig. 2, echo laser waveforms from vegetated surfaces include returns from the top of tree and ground, which enable us to measure the canopy height of tree.
The beam of the laser scanner is distributed into a certain swath of observation by scanning mechanism such as a galvanometer scanner or a polygon mirror scanner (see III –D). It enables us to observe the 3D information from not only the nadir of the spacecraft, but also off-nadir direction with uniform densities of footprints.
Air-borne laser scanner is widely used to capture 3D data of large areas, such as agricultural or forestry sites, urban areas, industrial plants, etc. On the other hand, space-borne laser scanner has not been realized yet. There was a space-borne laser altimeter observing the nadir direction, named the Geoscience Laser Altimeter (GLAS) which is the primary instrument of the Ice Cloud and Land Elevation Satellite (ICESat) of NASA . NASA is planning to launch ICESat-2 whose primary instrument is the Advanced Topographic Laser Altimeter System (ATLAS). The ATLAS laser is divided into 3 pairs of tracks, with about 3 km between pairs .
CONSIDERATIONS OF SPACE-BORNE LASER SCANNER
JAXA has been investigating a laser transmitting system for active light ranging missions in near future. In this study, we assume that this “space-based laser” is mounted to the GLS. Main specifications of the space-based laser are described below.
Three candidates of platform are considered as below.
• Earth observation satellite…600 km.
• International space station…400 km
• Super low altitude satellite…200 km
In considering a future mission using space-based laser, eye-safety of people on the ground must be considered. The safety standards of laser irradiation are prescribed by the International Electrotechnical Commission (IEC) .
These standards define the maximum permissible exposure (MPE) of a single laser pulese to the eyes is 50 mJ/m2 at λ =1064 nm.
In the case of space-borne laser sccaner, the situation that observer on the ground is viewing through a telescope (30 cm diameter) and all the laser light which enter the telescope aperture passes through his/her pupil (7 mm in diameter) is considered. In this case, the MPE is 27 μJ/m2 at λ =1064 nm, which is (30 / 0.7)2 times more strict criteria than the standard of the IEC.
Two candidates of scanning mechanisim, galvanometer scanner and polygon mirror scanner, are considered, and schematic diagrams of footprints distributed by galvanometer scanner and polygon mirror scanner are shown in Fig. 3. As shown in Fig.3, Density of footprints put by the galvanometer scanner is relatively high at the edges of the observation swath and low at the nadir. On the other hand, the polygon mirror scanner put footprint with uniform density. In this paper, the polygon mirror scanner is assumed to be a scanining mechanism of the GLS. More detiled trade-off study is needed until the final decision of the scanning mechanisim is made.
Assuming that the space-based laser is the light source of the GLS, maximum laser power is 15 W. According to limitation of eye safety, a laser pulse with 1 mJ must be expanded to 37 m2 area at the ground (atmospheric extinction is not considered). In this case, the total area of footprint per one second is about 56 ha (= 5.6 × 105 m2). Thus, about 1.7 Gha is observable per year, and it is corresponding to 3.3 % of all surface area of the Earth (overlaps between footprints are ignored.).
Required scanning rate frequency, maximum numbers of footprints per a scaning line, maximum observable swath, maximum observable frequency per year, and full scan angle for each interval of footprints.
|Hight of platform … 600 km|
|Interval of footprints.||100 m||250 m||330 m||500 m|
|Required scanning rate frequency||78 Hz||31 Hz||23 Hz||16 Hz|
|Maximum numbers of footprints per a scaning line||13||32||43||63|
|Maximum observable swath||1.3 km||8.0 km||14 km||31.5 km|
|Maximum observable frequency per year||0.4||2.4||4.2||9.3|
|Full scan angle||0.4 deg.||2.3 deg.||4.1 deg.||9.0 deg.|
|Hight of platform … 400 km|
|Interval of footprints||100 m||250 m||330 m||500 m|
|Required scanning rate frequency||77 Hz||31 Hz||23 Hz||15 Hz|
|Maximum numbers of footprints per a scaning line||13||32||43||67|
|Maximum observable swath||1.3 km||8.0 km||14 km||34 km|
|Maximum observable frequency per year||0.4||2.3||4.1||9.5|
|Full scan angle||0.2 deg.||1.1 deg.||2.1 deg.||4.8 deg.|
|Hight of platform … 200 km|
|Interval of footprints.||100 m||250 m||330 m||500 m|
|Required scanning rate frequency||76 Hz||30 Hz||23 Hz||15 Hz|
|Maximum numbers of footprints per a scaning line||13||33||43||67|
|Maximum observable swath||1.3 km||8.3 km||14 km||34 km|
|Maximum observable frequency per year||0.4||2.2||3.9||9.1|
|Full scan angle||0.1 deg.||0.8 deg.||1.4 deg.||3.2 deg.|
Placement of the Footprints
As a result of simple simlation for the GLS signal, we consider that about 15 mJ of laser pulse energy can be needed to obtain the data from which canopy height of tree can be estimated. Thus, the laser pule energy is assumed to be 15 mJ in this paper. According to limitation of eye safety, this laser pulse of 15 mJ is expanded to about 560 m2 area at the ground and the diameter of the footprint is about 27 m. And, maximum pulse rate frequency of the GLS is supposed to be 1000 Hz since the maximum laser power is 15 W, at this time.
As a function of interval between next footprints, required scanning rate frequency, maximum numbers of footprints per a scaning line, maximum observable swath, maximum observable frequency per year, and full scan angle are estimated, and they are summarized in Talbe 1. With smaller interval of footprints, higher frequency of the scanning rate are required and maximum observable swath is narrower, hence maximum observable frequency per year is lower than those of larger interval between next footprints.
In terms of the observation for terrestrial vegetation, 4 observations per a year is preferable . To make 4 observations per year, about 15 km observation swath is necessary (probability of cloud is not considered). The estimation is below.
Length of the Equator: 40075 km
Cycle number: 15 day-1
Recurrent days: 365/4 days
This swath is corresponding to full scan angles of 4.2, 2.1, and 1.4 degrees at 200, 400, and 600 km, respectively. Since required scan angles are relatively low, the waveform can be observed at the edge of the 15 km swath without much difficulty.
Thus when the interval of footprints is larger than 330 m, 4 observations per year can be achieved with the space-based laser which has been studied by JAXA.
For the optimazation of the footprints placement, we have disscussed with assumed users and further study will be required.
Simulation of Waveform from Vegetation Area to Estimate Canopy Height of Tree
To develop analysis algorithm for waveform of the GLS and to estimate observation accuracy of tree canopy height by the GLS, waveforms from some vegetation areas are simulated. Fig. 4 shows two examples of the waveform simulations, (a) from cedars, and (b) a zelkova (they are usual trees in Japan). Canopy height of the cedars and the zelkova are about 27 m and 11 m, respectively. 3D illustrations of the trees are in the charts. Thin cylinders above the trees represent laser footprints and diameters of the both cylinders are same. While a number of cedars were used for the waveform simulation, only one zelkova was generated in the fields of simulation since the zelkova has a large canopy area corresponding to the footprint.
Noise of the signals have not been simulated, thus the waveforms only represent intensity and observed time of photons reflected from the trees and the ground. In the case of cedars, there are interspaces between each cedars and intensity of return pulse from the ground is relatively high. On the other hand, a large part of photons are refracted by the zelkova and intensity of return pulse from the ground is low.
In the case of cedars which have acuminate shape, rising of the waveform is relatively slow. On the other hand, waveform of the zelkova shows relatively sharp initial rise. It is indicated that the ideal waveform represents vertical profile of projected area of trees. The initial rise of waveforms are fitted by the Gaussian function, and the half width at half maximum (HWHM) of the initial rise are 6.6 ns and 5.2 ns for the cedars and the zelkova, respectively. In these simulations, HWHM of the laser was 2 ns. Thus, contributions of the tree-canopy shapes to the HWHM are roughly estimated to be 6.3 and 4.8 for the cedars and the zelkova, respectively. Through the simulation and analysis as above, appropriate algorithm for the analysis of the GLS waveform has been studied.
There are various methods of analysis to estimate canopy height of tree from the waveforms obtained by the ICESat/GLAS , and a large part of the tree height error is likely to be caused by waveform analysis algorithm. Thus, the further study of the analysis algorithm of GLS waveform is important to improve the accuracy of the canopy height.
Other Application of the GLS
Space-borne laser scanner will provide other information on a wide range of Earth surface characteristics and processes. For example, the GLS will provide the global elevation map with high resolution which is potentially useful to understand and forecast damage that a stricken area may suffer. Observation of ice-sheets is also a candidate of the GLS mission, and the thickness of the ice-sheets in the polar region is considered to be one of the important observation objects to understand the global warming.
CONCLUSION AND FUTURE WORK
High-accuracy three-dimensional (3D) information of global area is very useful in various fields, such as global observations of tree height, elevation, and ice sheet. The Japan Aerospace Exploration Agency (JAXA) has started a conceptual study of earth observation missions with a space-borne laser scanner (GLS, as Global Laser Scanner) which is especially expected to observe canopy height of tree from the globe.
Assuming that the space-based laser which has been studied in JAXA is the light source, maximum laser power of the GLS is 15 W. We consider that about 15 mJ of laser pulse energy can be needed to obtain meaningful data. According to the limitation of eye safety, the laser pulse of 15 mJ is expanded to the footprint with about 27 m diameter at the ground. Maximum pulse rate frequency of the GLS is supposed to be 1000 Hz from the restriction of space-based laser. The total area of footprint per one second is roughly estimated to be 56 ha (= 5.6 × 105 m2). Thus, about 1.7 Gha is observable per year, and it is corresponding to 3.3 % of all surface area of the Earth.
Some candidates of the footprint placement are studied. When the interval of footprints size is larger than 330 m, 4 observations per year can be achieved.
For the optimization of the footprints placement, we have disscussed with assumed users and further study will be required.
To develop analysis algorithm for waveform of the GLS and estimate observation accuracy of tree canopy height by the GLS, waveforms from cedars and a zelkova are simulated. It is indicated that the ideal waveform represents vertical profile of projected area of trees. Since a large part of the tree height error is likely to be caused by waveform analysis algorithm in the case of ICESat/GLAS, the further study of the analysis algorithm of GLS waveform will provide important contribution to the expected accuracy of the canopy height of tree by the GLS.
The GLS will provide other information on Earth surface characteristics and processes, such as the global elevation map with high resolution and observation of ice-sheets.
The waveform simulator for the GLS was developed by collaborative study of Institute of Industrial Science, the University of Tokyo and JAXA. Especially, the authors would like to thank professor Haruo Sawada, assistant professor Takahiro Endo, Dr. Yoshito Sawada, Mr Shigeru Endo for developing of the simulator and waveform simulations.
F. G. Hall, K. Bergen, J. B. Blair, R. Dubayah, R. Houghton, G. Hurtt, J. Kellndorfer, M. Lefsky, J. Ranson, S. Saatchi, H. H. Shugart, and D. Wickland, “Characterizing 3D vegetation structure from space: Mission requirements”, Remote Sensing of Environment, vol. 115, pp. 2753–2775, 2011.Google Scholar
H. J. Zwally, B. Schutz, W. Abdalati, J. Abshire, C. Bentley, A. Brenner, J. Bufton, J. Dezio, D. Hancock, D. Harding, T. Herring, B. Minster, K. Quinn, S. Palm, J. Spinhirne, R. Thomas, “ICE Sat’s laser measurements of polar ice, atmosphere, ocean, and land”, Journal of Geodynamics, vol. 34, pp. 405–445, 2001.Google Scholar
A. W. Yu, M. A. Stephen, S. X. Li, G. B. Shaw, A. Seas, E. Dowdye, E. Troupaki, P. Liiva, D. Poulios, K. Mascetti, “Space Laser Transmitter Development for ICESat-2 Mission”, Proc. of SPIE Vol. 7578 757809–1, 2010.Google Scholar
K. Konoue, H. Kohata, S. Yamakawa, N. Igarashi, S. Imamura, “Development of Super Low Altitude Test Satellite (SLATS)”, 28th ISTS, 2011-n-06.Google Scholar
IEC825-2 and JIS C 6802Google Scholar
H. Duong, “Processing and Application of ICESat Large Footprint Full Waveform Laser Range Data”, Doctor thesis of the Delft University of Technology, Netherlands, 2010.Google Scholar