Memoirs

A semi empirical backscattering model of forest canopy covered by snow using SAR data

Description
A semi empirical backscattering model of forest canopy covered by snow using SAR data
Categories
Published
of 3
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  A SEMI EMPIRICAL BACKSCATTERING MODEL OF FOREST CANOPY COVERED BY SNOW USING SAR DATA zyxw li Nadir Arslan ), Jarkko Koskined2', Jouni Pulliainen'2', Martti Hallikained2' I) Nokia Research Center, P.O.Box 407,FIN-00045 NOKlA GROUP,Finland, E-mail: zyx rslan@Nokia.Com  (2) Helsinki University zyxwvuts f Technology, Laboratory of Space Technology, P.O.Box zyxwv 000 FIN-02015 HUT, Finland SEMI-EMPRICAL BACKSCATTERING MODEL ABSTRACT The multifrequency polarimetric EMISAR data at 5 3 GHz and ground-truth data such as forest and snow measurements were conducted in Northern Finland during EMAC95. The radar response to forest and snow characteristics, such as the stem volume and the snow water equivalent is analyzed. A semi-empirical backscattering model of forest canopy covered by snow which is a function of the forest stem volume and the snow water equivalent is developed. INTRODUCTION The EMAC'95 snow test sites are located along a straight line in northern Finland extending from the coast (near city of Oulu) to inland. The test sites are 9 zyxwvut m by 9 km squares and they consist of open areas (agricultural land) and forested areas (pine dominated forest). The analysis covers three snow test sites (1, 2 and 4). The land-use map of the test sites was obtained from the National Land Survey of Finland. The spatial resolution of the map is 25 m by 25 m. Ground truth data was collected along the center line of each test site by the Helsinki University of Technology and Finnish Environment Institute. Forest canopy data were provided by the Finnish Forest Research Institute. The ground-truth measurements include snow extent, snow depth, snow water equivalent, snow density profile, snow wetness profile, snow temperature, snow stratigraphy, snow grain size, land-use data (stem volume and land-use maps) and weather data. EMISAR data was collected on 22-23 March and 2-3 ,May 1995 for test sites 1,2 and 4[l]. The main problem with the forest canopy models is the complexity of the target. The more accurately the model includes the physical features of the target, the larger is the number of parameters needed. Empirical models typically have a substantially smaller number of parameters than theoretical models. Semi-empirical models can basically combine benefits of both modeling approaches. In this paper, a C-band semi-empirical backscattering model is presented for the forest-snow-ground system. The total backscattering coefficient of the forested snow- covered ground is divided into two (incoherent) contributions which are given as follows, zyx   0 20 r zyxwv ov +t U where Ov = the backscattering coefficient from forest canopy, z T = the-backscattering coefficient from ground, t 0 0 = the two way transmissivity of the canopy. The canopy transmissivity is determined as a function of stem volume by analyzing the change observed in the backscattering contribution of thte ground floor. The forest transmissivity at vertical polarization, 50 incidence angle, as function of frequencyf [GHz] and forest-stem volume V [m3 / ha] is given as follows[2], (2) where t f,Vhigh)= .42 + [l 0.42].exp{-0.028.f} which is frequency dependent transmissivity of a very dense forest, i.e. the saturation value of forest transmissivity at a certain frequency. 0 The backscattering coefficient of the forest canopy, 0 can be written in the following form[3],[4] and the two way canopy transmissivity : 2 (-2h, /cose)K, t =e where 0, volume backscattering coefficient, zy 0-7803-6359-01001 10.00 2000 IEEE 1904 (4)  8 zyxwvutsr   zyxwvutsrq ngle of incidence, zyxwvutsrq , = canopy extinction coefficient, h, = canopy (tree) height, We can rewrite (3) in terms of the stem volume as [ 3],[4]  zyxwvu OME RESULTS By using (2), we calculated the forest transmissivity depicted in Fig.1, at vertical polarization, 50 incidence angle, as function forest-stem volume V [m3 ha] at 5.3 GHz. where C is the empirical coefficient for canopy backscatter [ha / m3] zyxwv   is the empirical coefficient for canopy extinction [ha / m3] and Vis the forest volume [m3 ha]. The transmissivity behavior of forest canopy is assumed to be exponential as a function of stem volume according to (4) and (5). The values obtained for the coefficients above for zyxwv  3 GHz, VV polarization, 50 ncidence angle are C = 0.002 and U = zyxwvutsr .003. The backscattering coefficient from ground floor of the forest, bg , may be written as, (6) b , = 0; +,;os where Osp s backscattering coefficient of snowpack, 0, s backscattering coefficient of soil, 0 t, is transmissivity of snowpack. The backscattering coefficient from ground floor of the forest, Ob , s modeled empirically as follow [ 5],[ 61, o: , d~) uswe + zyxwvu   , (7) where a and b are constant coefficients and swe is snow water equivalent of dry snow (mm). The coefficients were determined by the least square sum fitting of zyxwvu 1) to the measurement data. The values obtained for a and b are 0.0228 and -34.744 respectively. 1 0.99 0.98 0.97 . 0.96 m 0.95 0.94 0.93 0.92 0.91 . .- i 0 20 40 60 80 100 Stem Volume mA3/ha) Fig. 1. The forest transmissivity at vertical polarization, 50 incidence angle, as function forest-stem volume V [m3 ha] at 5.3 GHz. Fig.2 shows the total backscattering coefficients and backscattering contributions from the forest canopy and the ground floor as function of stem volume at 5.3 GHz,VV polarization, 50 incidence angle. The solid lines depict the second-degree polynomial regression lines for the total backscattering coefficient, forest canopy and ground floor backscattering contributions. The canopy transmissivity (2) can be evaluated by comparing the 0; odeled by (5) with the second-degree polynomial regression curve determined from the data points. This is shown in Fig.3. The model and the second-degree polynomial regression curve are in good agreement with each other. 1905  Stem Volume zyxwvutsrqpo   mA3/ ha zyxwvutsr 1 20406080100 420 40 zyxwvutsrq 0 80 100 -10 zyxwvutsrqponml 51 m 50 A Ground Backscatter Fig.2. Backscattering contributions of covered by snow by the semi-empirical modeling approach. the forest canopy zyxwvu CONLUSIC,N The radar response to the forest campy covered by snow was studied by using EMISAR data. A semi-empirical backscattering model is presented flor the forest-snow-ground system. The backscattering coeffici,ent from forest canopy is modeled as a function of stem volume while the backscattering coefficient from ground floor, which is snow in this case, is modeled as a function of snow water equivalent. The forest canopy transmissivity used in this paper is modeled as a function of the frequency and :stem volume. The results showed that the semi-empirical model presented in this study is in good agreement with the experimental data. REFERENCB [ 1 M.Hallikainen, J.Koskinen, J.Praks, A.N.Arslan, H.Alasalmi, and P.Makkonen. Mapping of snow with airborne sensors in EMACP5, EMAC Final Workshop, ESA-ESTEC,Nordwijk, The Netherlands, 1997. [2] N.Kruopis, J.Praks, A.N.Arslan, H.Alasalmi,J.Koskinen, and M.Hallikainen, Passive microwave measurements of snow-covered forest areas in EMACPS, IEEE Trans. Geosci. Remote Sensing, vol. 37, pp. 2699-2705, November, 1999. [ 31 J.Pulliainen, K.Heiska, J.Hyylppa, and M.Hallikainen, Backscattering properties of boreal forests at the C- and X-band, IEEE Trans. Geosci Remote Sensing, vol. 32, pp. 1041-1050, September, 1994. [4] J.Pulliainen, P.Mikkela, and M.Hallikainen, Seasonal dynamics of C-band backscatter of boreal forests with applications to biomass and ,soil moisture estimation, IEEE Trans. Geosci. Remote Sensing, vol. 34, pp. 758- 770, May, 1996. [5] A.N. Arslan, J.Koskinen, J.Praks, H.Alasalmi, N Kruopis, and , M.Hallikainem, Retrieving snow water equivalence on C- and L-bancl SAR data for dry snow, Geoscience and Remote Sensing Symposium Proceedings ,1998. IGARSS '98. 1998 IEEE International Volume: 4, 19913 Page(s): 1870 -18'72. A.N. Arslan, J. Praks, J. Koskinen, and M.Hallikainen, An empirical model for retrieving water equivalent of dry snow from C-band SAR data, Geoscience and Remote Sensing Symposium, 1999. IGARSS'99 Proceedings. IEEE 1999 International Volume: 3 1999 Page(s): 1'789 -1791. [6] Fig.3. Forest canopy contribution 0;at zyxwvutsr  3 GHz, VV polarization. 19 6
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks