Southern California

Mapping and analysis of pre-fire fuels loading and burn intensity using pre-fire interferometric synthetic aperture radar data combined with burn intensity derived from post-fire multispectral imagery for the 2003 southern California fires.


In 2003 15 large fires consumed more than 750,000 acres in southern California. Coincidentally, a pre-fire, high-resolution interferometric radar dataset was collected in 2002/2003 by NOAA that covers most of the major burned areas from Santa Barbara to the Mexican border (~10 million acres). These data were collected with the GeoSAR interferometric synthetic aperture (IFSAR) radar system and are available from NOAA for analysis. Also, high-resolution multispectral imagery was flown over all major burned areas (~2 million acres) and used to produce burned area orthophotographs immediately after the fires. This post-fire imagery (along with BAER team fire intensity maps) is also available from ERSI / San Bernardino National Forest. In this project, we propose to: install and measure a limited number of vegetation field plots in unburned areas surrounding 3 major burned areas to augment existing national forest, CDF, FIA and NRI pre-fire inventory plot data; develop regression relationships between these ground plot vegetation data; use these regressions to map vegetation density, structure, and fuels throughout the burned and unburned areas of the 3 major fires; compare these vegetation density/structure maps with burned area assessment maps (vegetation mortality, soil severity) developed by the BAER program; conduct an analysis to determine if the IFSAR vegetation density/structure maps are well-correlated with the post-fire high-resolution multispectral orthophotography images and finally, make recommendations on whether or not it is desirable to process the entire NOAA IFSAR dataset to develop fuels maps throughout the 5 county area.

Credits and Responsiblities

Principal Investigators:

Dr. Hans-Erik Andersen
University of Washington

Robert J. McGaughey
PNW Research Station

Primary Federal Cooperator:

USDA Forest Service PNW Research Station

Robert J. McGaughey
University of Washington
School of Forest Resources
PO Box 352100
Seattle, WA 98195-2100
Phone: (206) 543-4713, FAX: (206) 685-3091, Email:

Ken Winterberger
Forest Inventory/Remote Sensing
Forest Inventory and Analysis Program
Anchorage Forestry Sciences Laboratory
Phone: (907) 743 9419, FAX: (907) 743 9482, Email:
Project contribution: FIA plot data and FIA plot analyses

Other Federal Cooperators:

San Bernardino National Forest

Sean Redar
Fire Management—Resource Information
1824 S. Commercenter Circle, San Bernardino, CA 92408
Phone: 909-382-2652, Email:
Project contribution: Inventory plot data, orthophotos, aerial photos

National Oceanic and Atmospheric Administration

Kirk Waters
NOAA Coastal Services Center
2234 South Hobson Avenue
Charleston, SC, 29405-2413
Phone: (843) 740-1227, Email:
Project contribution: IFSAR dataset

Non-Federal Collaborators:

EarthData International

Jim Reis
Phone: 301-948-8550, ext. 147, Email:
Project contribution: Additional X-band and P-band IFSAR data processing, ISTAR imagery, DEMs, technical consultation on IFSAR processing

California Department of Forestry and Fire Protection

Forest Resources Assessment Program
Robin Marose, Manager, Phone: (916) 227-2656, Email:
Mark Rosenberg, GIS/Fuels Mapping, Phone: (916)-227-2658, Email:
Project contribution: Consultations on recommended future use of IFSAR vegetation maps for CDF fuel mapping

ESRI, Inc.

Gerco Hoogeweg, Phone: (714) 372-9233, FAX: (714) 894 - 9699
Dr. Martha Sutula
7171 Fenwick Lane, Westminster, CA 92683
Project contribution: (via NOAA) IFSAR dataset

California Department of Conservation

Advanced Mapping Applications
Robert Yoha, ph: (916) 322-9307
801 K Street, Sacramento, CA 95814
Project contribution: Technical consultations on GeoSAR system

California State University - Fresno

Dr. Riadh Munjy, ph: (559) 278-4828
Project contribution: Technical consultations on GeoSAR system


In many parts of the West, there is severe fire danger as a result of high fuel loadings. Nowhere was this more evident in the 2003 fire season than in Southern California (SoCal), where over 650,000 acres burned in October in five counties (Ventura, Los Angeles, San Bernardino, Riverside, and San Diego). Reduction of fuels, particularly in wildland-urban-interface (WUI) areas is a national priority for the USDA Forest Service, and recently, the Healthy Forests Restoration Act of 2003 was signed into law to help speed the reduction of dangerous fuel loadings, with priority given to WUI areas. However, to better guide this effort, improved methods are needed to measure and map vegetation structure and associated fuel loadings. Additionally, improved methods for measuring fuel loading over landscapes are needed to build and validate fire behavior models. The emergence of a new generation of high-resolution remote sensing systems could potentially allow for more accurate and efficient estimation of fuels and fire behavior variables. With spatial resolutions in the sub-meter range, the spatial data provided by these sensors can support more detailed measurement of vegetation structure. The ability of active microwave (radar) airborne sensors to penetrate the canopy can significantly improve estimation of the quantity and distribution of vegetation structure, moisture regimes, and density.

This proposal is submitted under Joint Fire Sciences Program Solicitation 2004-1, Task 2 which solicits studies that use pre-fire fuels conditions, post-fire data, and fire behavior for 2003 fires, with emphasis on sites where pre-fire data are available.


We propose to use the extensive existing pre-fire and post-fire datasets to:

Evaluate the utility of the existing high-resolution GeoSAR dataset for vegetation structure mapping by developing high-resolution (1-3m range) pre-fire vegetation structure maps for 3 major 2003 SoCal fires.

Compare GeoSAR-derived vegetation structure maps with BAER post-fire burn assessment maps to determine if pre-fire IFSAR vegetation measurements correlate well with fire intensity.

Make recommendations, in consultation with CDF, regarding the desirability of using the existing GeoSAR dataset to improve fuels maps in SoCal.

Data Sets

Existing Data Sets

Pre-fire IFSAR

During 2002-2003, the National Oceanic and Atmospheric Administration, Coastal Services Center (NOAA) and the Southern California Wetland Recovery Project (SCWRP) contracted with EarthData International (EDI) to collect interferometric synthetic aperture (IFSAR) data over coastal watersheds from Point Conception to the US border with Mexico (Figure 1), which includes coverage of all the major fires in SoCal in 2003 in pre-burn condition. SCWRP is using this IFSAR dataset to develop GIS-based tools for prioritizing wetland restoration and conservation options. Analyses of riparian areas are being done across the region to identify areas with high ecological value and to examine the costs and benefits of using land-use and land-cover data collected at different spatial scales to map riparian vegetation. The project is also developing conceptual models that examine the habitat, hydrology, and biogeochemistry functions of wetlands within their landscape context. The SCWRP is a multi-agency effort within California and is led by the California Coastal Conservancy. Although the dataset was collected for analyses of riparian and wetland conditions, it covers the entire landscape and provides a unique opportunity to collaborate with SCWRP to develop region-wide approaches for high-resolution vegetation structure maps.

The entire IFSAR dataset has been made available for the proposed project by NOAA, SCWRP and EDI. It is the first of its kind ever collected over a large area in the United States and offers the opportunity to develop pre-fire vegetation structure, density, and biomass measurements and maps at unprecedented resolution (1-3 meters), and for the first time, with direct measurement of vertical vegetation structure. EDI's GeoSAR system collects 0.5 m resolution X-band IFSAR data at a rate of 8,800 sq km per hour, offering for the first time the ability to map and monitor regional-scale areas in a cost-effective manner. In the near future this type of IFSAR data will be available over large areas.

Post-fire IFSAR

Immediately following the 2003 SoCal fires, EDI was also contracted to collect high-resolution digital infrared imagery (0.5 m) over the major fires (Figure 1). This imagery has also been made available for this proposed study. It was used to produce post-fire orthophotography and digital elevation models (DEMs) of unprecedented resolution and accuracy (3 m resolution, 0.5 m accuracy, respectively).

FIA ground plot data

Within and surrounding the fires ground vegetation data have been collected on a systematic grid (5km x 5km) by the PNW Research Station, Forest Inventory and Analysis (FIA) Program. Although sparse, these plots are georeferenced and available to be combined with additional ground plot data that will be collected specifically for the proposed project. Note that Ken Winterberger, PNW FIA Program, is a collaborator on the proposed project, thus making plot georeference data available in the proposed analysis.

BAER Program post-fire assessment maps

Burn Area Emergency Response teams have produced post-fire assessment maps for vegetation mortality, soil burn severity, and watershed response. These maps are available ( and will be compared to pre-fire IFSAR vegetation structure.

CDF fire fuels maps

The California Department of Forestry and Fire Protection (CDF) has produced fire fuels maps using LANDSAT imagery and FIA plot data. These maps will be used in the final stages of the proposed study when recommendations are developed in collaboration with CDF on whether or not it is desirable to integrate the IFSAR vegetation maps into future regional fuel maps.

Figure 1. Extents of IFSAR data coverage (white outline) and ISTAR imagery, orthophotos, and DEMS (red outlines) available for proposed vegetation mapping.

Urgency of this project:

To capture the maximum value out of this existing IFSAR dataset, it is important that this study commence as soon as possible. First, it is very likely that additional large fires will burn in the area in 2004, reducing the area into which suitable unburned ground plots can be located around the 2003 burned areas. Second, as each year passes, the unburned vegetation structure will continue to change in areas where ground plots are to be installed, thus making it more difficult to develop relationships with the existing IFSAR dataset that was collected in winter 2002-2003. Finally, another IFSAR provider (Intermap) has initiated an aggressive program to collect similar IFSAR data over the entire lower 48 states. In 2004, Intermap plans to collect IFSAR data for over 1 million sq km in Mississippi, California, Nevada, and West Virginia. (The existing SCWRP IFSAR dataset covers only 40,000 sq km). What is learned in the analysis proposed in this study will answer many questions about the utility of new data as it becomes available and recommended specifications for future data collection.

Remote Sensing data:

The IFSAR remote sensing technology

There is tremendous potential for the use of IFSAR (Interferometric Synthetic Aperture Radar) to estimate the distribution of canopy fuels over large areas at relatively low cost. It is believed that the three-dimensional forest structure information derived from interferometric radar data could be directly related to density and spatial distribution of canopy biomass and bulk density.

SAR (Synthetic Aperature Radar) is an active sensing technology that emits and records the reflection of microwave radio energy. The information content of radar data in forested terrain varies depending upon the wavelength (l) of the transmitted pulses – energy with short wavelengths (l ? 1 cm) is reflected from the canopy surface while radar energy with longer wavelengths (l ? 1 m) penetrates the foliage in the canopy and reflects from tree trunks and the terrain surface. Another characteristic of microwave remote sensing, in contrast to optical remote sensing, is the capability to penetrate cloud and smoke cover. The resulting image represents the intensity of the radar backscatter throughout the illuminated region. Because the reflection of the radar signal is dependent upon the dielectric properties of the scattering elements within the resolution cell, SAR can also be used to measure soil moisture and canopy water content. While previous studies have shown that SAR backscatter amplitude data can be used to estimate forest biomass (Hussin et al., 1991), it has been noted that the biomass saturation limits for even long-wavelength SAR systems (~ 150 tons/ha) are too low to reach levels present in temperate closed forests (~ 300 tons/ha) (Mette et al., 2003).

The availability of three-dimensional interferometric radar (IFSAR) data in recent years has the potential to significantly expand the applicability of radar analysis for forest structure analysis. Radar interferometry uses the difference in phase, or phase shift, between two radar images acquired from slightly different locations to acquire information relating to the elevation angle to an imaged point, which is used in conjunction with the range information to determine the three-dimensional location of this imaged point (Hagberg et al, 1995). Varying the wavelength of the emitted energy will allow collection of different three-dimensional structure data – sensors emitting pulses with short wavelength (i.e. X-band IFSAR with l = 3 cm) can measure the surface structure of the forest canopy, while sensors with longer wavelengths (P-band IFSAR with l = 72 cm) will generate a surface corresponding to the terrain elevation (Hofmann et al, 1999; Schwäbisch and Moreira, 1999). Accuracies of these systems also vary with wavelength; X-band interferometric radar data can have a vertical accuracy of 1-2 m, while P-band data has a vertical accuracy of 3-5 m (for GeoSAR system).

Past research has shown that polarimetric interferometric radar (PolInSAR) data acquired from single-frequency systems with wavelengths in the intermediate range (C- and L-band) can be used to extract information relating to the depth of various vegetation layers, the density of the scattering medium (related to biomass), and the elevation of the terrain surface (Treuhaft et al, 1996; Cloude and Papathanassiou, 1998). While many of these studies assumed an (admittedly simplistic) homogeneous density for the vegetation layer to reduce the number of parameters in the model, they have established the theoretical basis for more complex, and realistic, inferential approaches to the estimation of canopy density characteristics from IFSAR data. These authors have also noted that accuracy in the estimation of vegetation density and canopy characteristics would be expected to improve significantly through the analysis of multifrequency IFSAR data, such are available for this study.

The GeoSAR multifrequency IFSAR system

Multi-frequency IFSAR (X- and P-band) systems can provide canopy- and terrain-level elevation models as standard deliverable products. The GeoSAR system is a multifrequency IFSAR system mounted on a Gulfstream II jet operating from a flying altitude of 15,000 – 30,000 feet which acquires both X-band and P-band data in a single-pass mode at a rate of 8,800 km2 per hour (see Figure 2). System parameters for the GeoSAR system are given in Table 1.

Figure 2. GeoSAR multifrequency IFSAR system.

Center Frequency9.755 GHz350 MHz
Bandwidth80/160 MHz80/160 MHz
Peak Transmit Power8kW4kW
PolarizationV VHH, HV
Swath Width20 km20 km

The flight path is configured such that each point on the ground is imaged four times, from four different look angles. Co-polarized (HH, or horizontal transmit – horizontal receive) and cross-polarization (HV, or horizontal transmit-vertical receive) information is available at the P-band, while X-band data is acquired in co-polarized (VV) mode.

The difference of the canopy elevation (X-band) and underlying terrain elevation (P-band) yields a canopy height model that represents a spatially-explicit description of canopy structure (i.e. volume, height, biomass, etc.) over a given area of forest. An example of a IFSAR-based canopy height model over a conifer forest in Capitol State Forest, WA, along with an orthophoto of the same areas, is shown in Figures 3 & 4. The use of multi-frequency (X-band and P-band) IFSAR systems for forest mapping has emerged relatively recently, where research efforts have largely focused on improving forest type classification (Hofmann et al., 1999; Dutra et al. 2002; Mura et al., 2001).

Figure 3. Orthophoto of 5 km2 area within Capitol State Forest, WA

Figure 4. IFSAR-based canopy height model, 5 km2 area within Capitol State Forest, WA

Figure 5. Comparison of profiles generated from IFSAR, LIDAR, and photogrammetry (location of profile shown in Figure 3) (from Andersen et al., 2003)

Ultimately, the value of IFSAR as a source of data for fuels mapping will be directly dependent upon the accuracy of the system in measuring both the canopy structure and underlying terrain surface. Several studies have evaluated the accuracy of IFSAR-based (X-band) digital terrain models, generally finding that IFSAR DTMs have accuracies of 0.6 – 1.5 meters (Mercer, 2001; Norheim et al., 2002). A recent study has shown that an IFSAR canopy surface model (CSM) generated from IFSAR (X-band) data provided an accurate model of the canopy surface structure when compared to photogrammetric measurements acquired from large-scale photographs (Andersen et al., 2003). Figure 5 shows a comparison of profiles generated from an IFSAR-derived (X-band) canopy surface model, IFSAR (P-band) digital terrain model, and a LIDAR-derived digital terrain model within a dense conifer forest area. Comparison of LIDAR terrain model to the IFSAR (P-band) terrain model also indicated that the IFSAR-based (P-band) terrain surface provides an accurate measurement of terrain elevation even under dense forest canopy (see Figure 5).

Methods and Study Areas

Estimating canopy fuels from multifrequency polarimetric IFSAR data

Previous ecological research has shown that the relationship between forest canopy height and biomass is governed by the basic law of proportional growth, expressed in the allometric equation (West, 1997):


This relationship has been used in previous studies to relate LIDAR-based canopy height measurements to stand biomass (Lefsky et al., 2001). A more recent study has used allometric equations to relate IFSAR-based (L-band) canopy height measurements to biomass in Germany (Mette et al. 2003). While this study has shown that these relationships are stable across a range of thinning regimes, it was noted that biomass levels in more complex, natural forests will also depend upon stand density and vertical forest structure. It is expected that other components of the IFSAR data, including polarimetric radar backscatter amplitude, interferometric coherence, and differential geometric information provided by multiple looks will provide valuable information in refining biomass estimates in more complex forest types. As co-polarized radar energy (HH) tends to penetrate vegetation more than cross-polarized energy (HV), the comparison of backscatter amplitudes from multi-polarized P-band data should aide in quantifying structural characteristics within natural stands. Traditional (orthorectified) aerial photography (and possibly satellite imagery) could provide an additional source of information in the analysis of stand density and composition.

As biomass is related directly to crown bulk density, the theory described above can be used to estimate crown bulk density from the canopy dimensions, interferometric parameters, and backscatter information provided by multifrequency IFSAR data. A JFSP-funded research project at the University of Washington (JFSP project 01-1-4-07) is currently investigating the use of IFSAR (X- and P-band) data for measurement of critical crown fire behavior variables, including crown bulk density, canopy height, crown base height, and canopy cover, along with topographic variables such as aspect, slope, and elevation (Andersen et al., 2004). This study is comparing IFSAR-based measurements of canopy height and stand structure to field-based estimates of canopy fuels developed from tree list data acquired at 125 georeferenced field plots located within Capitol State Forest in western Washington. This research has been supported by the concurrent development of a scientific visualization system designed to graphically explore multidimensional IFSAR and LIDAR datasets (McGaughey et al., 2003). An example of the graphical display of multifrequency IFSAR data within this study area is shown in Figure 6.

Figure 6. Three-dimensional graphical visualization (top) of X-band IFSAR elevation measurements (color coded by height) overlaid on P-band digital terrain model and corresponding orthophoto (bottom).

Estimates of canopy fuels at each field plot were generated using the approach developed by Scott and Reinhardt (2001) for use in the Fire and Fuels Extension to the Forest Vegetation Simulator (FVS) (Reinhardt and Crookston, 2003). The relationships between IFSAR-derived estimates of vegetation biomass and crown bulk density were related to field-based estimates of canopy fuels using multiple regression analysis. The predictive models developed through the exploratory regression analysis are validated using independent data (if available) or cross-validation procedures. Preliminary results indicate that this approach can be used to efficiently estimate canopy fuel distributions from IFSAR (X-band/P-band) data within Pacific Northwest conifer forests (Andersen et al., 2004).

Selection of sample sites within burned areas of Southern California

In order to assess the capabilities of the IFSAR technology for quantifying canopy fuels in fire-prone areas of southern California, we will select at least three major fires that burned in 2003 falling within the extent of the GeoSAR IFSAR dataset collected by Earthdata International through the Southern California Wetlands Recovery Project funded by NOAA. In order to allow for generalization of the results across the full range of the ecotypes within this region, these fires will be selected to include a broad range of vegetation types and pre-fire forest conditions. Sample areas (approximately 10-15 km2 in size) will then be selected within each selected fire, and will be chosen to include a range of burn severities (from unburned to totally consumed) and pre-fire vegetation structures (open/shrub to dense forest canopy).

Generation of IFSAR-based (P-band) digital terrain models for study sites

In order to minimize interference with both civilian and military radio communications within this region of southern California, the signal of the P-band radar signal waveform had to be “notched” – leading to a decrease in the bandwidth and a degradation of the interferometric resolution. For this reason, the IFSAR provider (Earthdata) processed only a small percentage of the P-band data acquired over the total area. Because the P-band IFSAR data are critical to the determination of terrain elevation under forest canopy (and therefore forest canopy height), it is proposed that Earthdata process the P-band data and provide standard IFSAR deliverables (strip elevations, HV/HH backscatter amplitudes, coherence, error estimates) for the selected sample areas. The accuracy of the P-band terrain models in post-burn open areas will be assessed through comparison to a DTM generated via digital photogrammetry from ISTAR digital imagery with vertical accuracy of 50 centimeters.

Estimation of canopy fuels within three selected sample areas

At least thirty field plots will be needed to enable calculation of fuel estimates from the plot-level tree list data. It is expected that 20 – 30 field plots will be established in unburned areas to augment the existing vegetation plot data provided by the Forest Inventory and Analysis (FIA) program and the San Bernardino National Forest inventory program. The location of the plots will be chosen to represent the full range of vegetation types and stand structures present within the study area. These field plots will be georeferenced to a 1-meter accuracy using differentially corrected GPS and surveying equipment (e.g. Criterion laser system). Field-based estimates of canopy fuels at each plot location will be developed using the techniques developed by Scott and Reinhardt (2001) and the FERA team.

A regression analysis (with validation component -- described above) will be used to establish the mathematical model(s) relating the IFSAR-derived vegetation parameters to the field-based fuel estimates. Probable predictor variables used in the regression analysis include an IFSAR-derived canopy cover estimate, quantile-based measurements of X-band heights, interferometric coherence, and polarimetric radar backscatter amplitude. These regression models will then be used to estimate the canopy fuel characteristics over the entire extents of the three major selected fires.

Comparison of pre-fire canopy fuel distributions to BAER post-fire assessment maps

The availability of the GeoSAR data set, acquired before the fires of 2003, offers a unique opportunity to investigate the relationship between the pre-fire canopy fuel loading, topographic conditions, and the intensity of fires that burned in 2003. Although fire severity is also heavily influenced by meteorological conditions, it is expected that fire intensity is highly correlated with the spatial distribution of canopy fuels and the topographic expression of the landscape. In order to quantify these relationships, the canopy fuel maps and topographic information generated from the IFSAR dataset for the three selected major fires will be compared to the post-fire burn severity maps developed by the Burned Area Emergency Response (BAER) program. These maps are generated using a variety of geographic data, including Burned Area Reflectance Classification (BARC) imagery, and field data. Comparisons will be made between the BAER vegetation mortality, soil burn severity, and watershed response maps and the IFSAR canopy fuel maps.

Recommendations for future work

The results of this analysis could be used to support a variety of follow-up projects. If the results of the vegetation fuels analysis within the selected study areas are encouraging, the entire existing GeoSAR P-band dataset could be processed and a canopy fuel map could be generated covering the entire five-county region of southern California. The high-resolution forest structure information provided by the GeoSAR dataset could be used to augment the fuel maps currently produced by the California Department of Forestry using low-resolution LANDSAT satellite imagery and FIA ground plot data. Given the future availability of IFSAR data over the entire United States collected through the NextMapUSA program, it is expected that this data could be incorporated into the FIA inventory to allow for generation of high-resolution maps of vegetation structure and canopy fuels over entire geographic regions.

It is also expected that the canopy fuel maps produced from high-resolution IFSAR systems could provide a valuable input to fire behavior models, such as FARSITE, which could support a detailed, simulation-based analysis of fire spread and intensity within the areas that burned in 2003 (Finney, 1998).


Pre-fire vegetation structure and fuel maps for three major fire areas

Evaluation of utility of GeoSAR data for mapping vegetation structure and fuels

Recommendatations for future use of existing GeoSAR dataset

Methodology for processing IFSAR data for vegetation structure mapping

Visualization software for exploring and displaying 3-D IFSAR datasets

Science Delivery and Applications

Present workshop to interested SCCWRP/NF/CDF personnel and fire behavior scientists on application of high resolution IFSAR vegetation maps for three major SoCal fires

Provide IFSAR processing and visualization software to SCCWRP/NF/CDF and fire behavior scientists

Publish results in scientific journals

Present results at national scientific conferences


Andersen, H.-E., W.W. Carson, R.J. McGaughey, S.E. Reutebuch, and G.F.Schreuder. 2004. Estimating crown fire behavior variables in a Pacific Northwest conifer forest using multifrequency polarimetric InSAR (PolInSAR). (submitted to ISPRS Congress, Istanbul, 2004).

Andersen, H.-E., R.J. McGaughey, W.W. Carson, S.E. Reutebuch, B.Mercer, and J. Allan. 2003. A comparison of forest canopy models derived from LIDAR and InSAR data in a Pacific Northwest conifer forest. International Archives of Photogrammetry and Remote Sensing. Volume XXXIV, part 3/W13. Dresden, Germany.

Cloude, S.R. and K.P. Papathanassiou. 1998. Polarimetric SAR Interferometry. IEEE Transactions on Geoscience and Remote Sensing 36(4): 1551-1565.

Dutra, L., M. Elmiro, C. Freitas, J.R. Santos, J.C. Mura, and B. Filho, 2002. The use of multi-frequency interferometric products to improve SAR imagery interpretability and classification by image fusion. In: Anais do III Workshop em Tratamento de Imagens, Junho 2002, Belo Horizonte, Minas Gerais, Brazil.

Finney, M.A. 1998. FARSITE: Fire area simulator - model development and evaluation. USDA Forest Service, General Technical Report RMRS-RP-4.

Hagberg, J., L. Ulander, and J. Askne. 1995. Repeat-pass SAR interferometry over forested terrain. IEEE Transactions on Geoscience and Remote Sensing 33(2): 331-340.

Hoffmann, C., M. Schwäbisch, S. Och, C. Wimmer, and J. Moreira. Multipath P-band interferometry – first results. Proceedings of the Fourth International Airborne Remote Sensing Conference and Exhibition/21st Canadian Symposium on Remote Sensing. Ottawa, Ontario, Canada, June 21-24, 1999.

Hussin, Y.A., R.M. Reich, and R.M. Hoffer. 1991. Estimating slash pine biomass using radar backscatter. IEEE Transactions on Geoscience and Remote Sensing 29(3):427-431.

Lefsky, M., W. Cohen, D. Harding, G. Parker, S. Acker, and S. Gower. 2001. Lidar remote sensing of aboveground biomass in three biomes. International Archives of Photogrammetry and Remote Sensing, XXXIV-3/W4:150-160.

McGaughey, R. J. and W.W. Carson. 2003. Fusing LIDAR data, photographs, and other data using 2D and 3D visualization techniques. In: Proceedings of Terrain Data: Applications and Visualization – Making the Connection, October 28-30, 2003; Charleston, South Carolina: Bethesda, MD: American Society for Photogrammetry and Remote Sensing. pp. 16-24.

Mercer, B., 2001. Comparing LIDAR and InSAR: What can you expect? In: Fritsch/Spiller, eds. Proceedings of the Photogrammetric Week 2001, Stuttgart, Germany. pp. 2-10.

Mette, T. K. P. Papathanassiou, I. Hajnsek, and R. Zimmermann. Forest biomass estimation using polarimetric SAR interferometry. 2003. Proceedings of the PolInSAR Conference, Frascati, Italy, Jan. 2003.

Mura, J., L. Bins, F. Gama, C. Freitas, J. Santos, and L. Dutra, 2001. Identification of the tropical forest in Brazilian Amazon based on the DEM difference from P- and X-band interferometric data. In: Proceedings of the Geoscience and Remote Sensing Symposium 2001, IGARSS ’01, Sydney, Australia. IEEE, vol. 2. pp. 789-791.

Norheim, R., V. Queija, and R. Haugerud. 2002. Comparison of LIDAR and InSAR DEMs with dense ground control. In: Proceedings of the Environmental Systems Research Institute 2002 User Conference, San Diego, USA.

Reinhardt, E.D. and N. L. Crookston. 2003. The Fire and Fuels Extension to the Forest Vegetation Simulator. General Technical Report RMRS-GTR-116. Ogden, UT: USDA Forest Service, Rocky Mountain Research Station.

Schwäbisch, M. and J. Moreira, 1999. The high resolution airborne interferometric SAR AeS-1. In: Proceedings of the Fourth International Airborne Remote Sensing Conference and Exhibition/ 21st Canadian Symposium on Remote Sensing, Ottawa, Ontario, Canada.

Scott, J.H. and E.D. Reinhardt. 2001. Assessing crown fire potential by linking models of surface and crown fire behavior. Research Paper RMRS-RP-29. Fort Collins CO: USDA Forest Service, Rocky Mountain Research Station.

Treuhaft, R.N., S.N. Madsen, M. Moghaddam, and J.J. van Zyl. 1996. Vegetation characteristics and underlying topography from interferometric radar. Radio Science 31(6): 1449-1485.

West, G.B., J.H. Brown, and B.J. Enquist. 1997. A general model for the origin of allometric scaling laws in biology. Science 476:122-126.