18
ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION IMAGERY FOR CROP ANALYSIS Sean Griffin, Remote Sensing Crop Analyst Ed Kunz, Geospatial Program Manager ASRC Management Services 6303 Ivy Lane, Suite 130 Greenbelt, MD 20770 [email protected] [email protected] ABSTRACT When developing estimates of crop yield typically a convergence of evidence including weather patterns (rain and heat) ground truthing (actual observations) and remotely sensed data are used. There are numerous areas in the world wherein one or several of the aforementioned areas of evidence are not available. The paper will focus on ways in which remotely sensed data (satellite imagery of medium and high resolution) is being used to determine crop yield at the field level (several acres) in place of ground observations which due to conditions could not be obtained. The paper will describe the satellite imagery used, detail processes and procedures utilized to fuse the data, explain how the data was analyzed and discuss the results obtained. INTRODUCTION The United States Department of Agriculture (USDA) Foreign Agricultural Service (FAS) is primarily responsible for overseas agricultural activities such as U.S. agricultural promotion and market development, international trade agreements, in addition to data collection and analysis (USDA FAS, 2008). The International Production Assessment Division (IPAD) within FAS is responsible for producing the objective, timely, and accurate assessments of conditions affecting world food security and providing global agricultural production forecasts. IPAD has a rich history in applying Geographic Information Systems (GIS) and Remote Sensing data analysis to crop production forecasting. IPAD essentially adopted the CADRE database system that was specifically designed for global agricultural monitoring by the Large Area Crop Inventory Experiment (LACIE) authorized in 1974, and later by AgRISTARS in 1980 (Colwell, 1983; Boatwright and Whitefield, 1986; Tingley, 1988). The CADRE database has been updated with a variety of observation data and processing models that are now displayed in public domain through a web interface called Crop Explorer. (http://www.pecad.fas.usda.gov/cropexplorer/ ). IPAD analysts are responsible for large regional coverage and typically focus on important agricultural commodity producers. Regions can cover spatial extents of several countries to entire continents in which 75 agricultural attaché and foreign national posts are made available to provide field travel, ground observations, and official country reports (Figure 1). However, little to no ground truth agricultural data are available for large agricultural importers and food insecure nations such as Iraq, especially given the limitations of travel in a country that was characterized by a military presence and personal safety concerns. In October 2007, FAS partnered with the Department of Defense (DOD) National Geospatial-Intelligence Agency (NGA) to establish the Iraq Operational Agricultural Monitoring Project. The purpose of the project was to develop an operational methodology that could function as an agricultural early warning system and provide timely and accurate assessments of Iraq’s field crops to U.S. and Iraq government officials, and non-governmental organizations involved with reconstruction efforts. Importantly, the project was developed to provide agricultural information without endangering ground forces or adding to their work load. This effort was also initiated to enhance FAS’s existing analysis capabilities by providing more direct analysis support to the rest of the region, as well as transfer these capabilities to other regions of the world where the lack of agricultural information resources is common.

DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION IMAGERY FOR CROP ANALYSIS

Sean Griffin, Remote Sensing Crop Analyst Ed Kunz, Geospatial Program Manager

ASRC Management Services 6303 Ivy Lane, Suite 130 Greenbelt, MD 20770

[email protected] [email protected]

ABSTRACT

When developing estimates of crop yield typically a convergence of evidence including weather patterns (rain and heat) ground truthing (actual observations) and remotely sensed data are used. There are numerous areas in the world wherein one or several of the aforementioned areas of evidence are not available. The paper will focus on ways in which remotely sensed data (satellite imagery of medium and high resolution) is being used to determine crop yield at the field level (several acres) in place of ground observations which due to conditions could not be obtained. The paper will describe the satellite imagery used, detail processes and procedures utilized to fuse the data, explain how the data was analyzed and discuss the results obtained.

INTRODUCTION

The United States Department of Agriculture (USDA) Foreign Agricultural Service (FAS) is primarily responsible for overseas agricultural activities such as U.S. agricultural promotion and market development, international trade agreements, in addition to data collection and analysis (USDA FAS, 2008). The International Production Assessment Division (IPAD) within FAS is responsible for producing the objective, timely, and accurate assessments of conditions affecting world food security and providing global agricultural production forecasts. IPAD has a rich history in applying Geographic Information Systems (GIS) and Remote Sensing data analysis to crop production forecasting. IPAD essentially adopted the CADRE database system that was specifically designed for global agricultural monitoring by the Large Area Crop Inventory Experiment (LACIE) authorized in 1974, and later by AgRISTARS in 1980 (Colwell, 1983; Boatwright and Whitefield, 1986; Tingley, 1988). The CADRE database has been updated with a variety of observation data and processing models that are now displayed in public domain through a web interface called Crop Explorer. (http://www.pecad.fas.usda.gov/cropexplorer/).

IPAD analysts are responsible for large regional coverage and typically focus on important agricultural commodity producers. Regions can cover spatial extents of several countries to entire continents in which 75 agricultural attaché and foreign national posts are made available to provide field travel, ground observations, and official country reports (Figure 1). However, little to no ground truth agricultural data are available for large agricultural importers and food insecure nations such as Iraq, especially given the limitations of travel in a country that was characterized by a military presence and personal safety concerns.

In October 2007, FAS partnered with the Department of Defense (DOD) National Geospatial-Intelligence Agency (NGA) to establish the Iraq Operational Agricultural Monitoring Project. The purpose of the project was to develop an operational methodology that could function as an agricultural early warning system and provide timely and accurate assessments of Iraq’s field crops to U.S. and Iraq government officials, and non-governmental organizations involved with reconstruction efforts. Importantly, the project was developed to provide agricultural information without endangering ground forces or adding to their work load. This effort was also initiated to enhance FAS’s existing analysis capabilities by providing more direct analysis support to the rest of the region, as well as transfer these capabilities to other regions of the world where the lack of agricultural information resources is common.

Page 2: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

STUDY AREA

Phase I of the project was initiated during the 2007/08 winter grains season and concentrated on the predominantly rainfed region of northern Iraq (Figure 2). This region was chosen due to historical variability and dependence on agro-meteorological conditions, as well as the regions potential for producing up to 40% of the total national wheat crop and up to 60% of the total barley output (Figure 3). Any negative impact on this region can adversely affect the food security of the entire country. Additionally, Iraq was chosen as a pilot study area because it provided a generalized representation of rainfed crop management and practices in similar food insecure Middle East nations, particularly the Fertile Crescent.

METHODOLOGY

The International Production Assessment Division (IPAD) applies a “convergence of evidence” methodology to agricultural production assessments and forecasts. This all source methodology integrates both qualitative and quantitative datasets of various spatial and temporal scales in order to cross-validate the multiple sources and scrutinize forecast assumptions. Datasets are in the form of

• Agro-meteorological time-series • Land observation remote sensing • Economic analysis reports • Crop calendars • News services • Direct ground observations. The final output of the analysis is a sub-national level crop production forecast that will ultimately enable food

aid agencies to determine impacts on supply and respond in a timely manner thus ensuring national food security and stability.

The utilization of satellite remote sensing enables multi-temporal analysis over large spatial extents. The strength of this methodology in relation to agricultural monitoring relies heavily upon seasonal comparisons and anomalous observations. In the past, IPAD applied a hierarchical remote sensing analysis approach that relied on coarser resolution sensors (250-meter + ground resolution) such as AVHRR and SPOT-VEG to discriminate vegetation anomalies over a large regional extent while deriving field level observations from moderate-resolution 30-meter Landsat-5 TM/Landsat-7 ETM+. However, due to a combination of crop season cloud cover and a 16-day repeat period, the reality of Landsat as an operational agricultural monitoring sensor was never met* (Tetrault, 2006). The scan line anomaly that appeared in the Landsat-7 ETM+ sensor and the inability of Landsat -5 TM due to age and mechanical issues to reliably collect OCONUS imagery forced IPAD to consider other alternatives. Consequently, the Indian Remote Sensing (IRS) AWiFS sensor was chosen to replace Landsat. The AWiFS IRS P-6 sensor boasts a nominal 56-meter spatial resolution with 4 spectral bands ranging from 0.52 to 1.70 microns. Furthermore, a 740km swath width combined with a potential revisit period of 5-days met the operational criteria of this project (ISRO, 2008). In lieu of non-existent ground-truthing data, high resolution Digital Globe Quickbird imagery was chosen as a source of image-based ground-truthing. Quickbird provides multispectral senor capabilities at an approximate spatial resolution of 2.5-meters (Digital Globe, 2008).

The winter grains season in Iraq begins with seed sowing from mid-October to mid-December, coinciding with the rainy season. Crop emergence and maturity occurs from mid-January to early-April, concluding with crop harvest from late-April to early-June. (Figure 4). Phase I of the project began by collecting imagery the first week of November 2007 and was scheduled for completion by the end of May 2008. A total of 17 areas of interest (AOI) within the rainfed agricultural zone were chosen for Quickbird image acquisition; 4 images per AOI were acquired throughout different phases of the crop season (Figure 5). AWiFS IRS P-6 provided country-wide coverage and was collected as frequent as possible, typically providing monthly cloud-free color and vegetation index composites (Figures 6). Daily MODIS imagery from the MODIS Rapid Response System and MODIS Normalized Difference Vegetation Index (NDVI) products were provided through the Crop Explorer interface (Figure 7). All image data were georectified and standardized to surface reflectance values before being used in analysis (Figure 8).

* The recent USGS release of +35-year Landsat archive provides an important resource for determining and documenting multi-temporal agricultural events in addition to change analysis studies.

Page 3: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Phase I start of season agricultural monitoring entailed querying the +30 year FAS agro-meteorological time series database provided in CADRE, in addition to NOAA Climate Prediction Center (CPC) datasets available in GIS-ready format. The aforementioned datasets were queried to provide an understanding of the current conditions that will ultimately impact future crop production and to discern anomalous crop years. In the case of rainfed winter grains, the most important variables during the crop establishment stage are temperature and precipitation. Start of season analysis included the use of Quickbird imagery to provide contextual evidence of field preparation. Also, daily MODIS imagery was used to confirm cloud cover during recorded precipitation events and MODIS NDVI time series data were used to provide a 10-year regional perspective of normal versus anomalous crop conditions. MODIS NDVI anomalies essentially served as a geographical guide to selecting areas of analytical concentration. The next step in the hierarchical analysis was to derive crop health assessments using moderate resolution AWiFS. Finally, the crop health assessments were either confirmed or denied using Quickbird as the validation source. Archived Landsat imagery coarsened to the spatial scale of AWiFS (30-meters to 56-meters) also served as a useful determinant of “better or worse” than previous years.

Agricultural production statistics were derived by calculating crop area and yield using information derived from multiple sensors at different phases of the crop season and the input of time-series crop statistics. Since the input data and final output are provincial level crop statistics, it is assumed that the majority of rainfed crops are comprised only of wheat and barley, and that yield remains constant throughout the province. Crop area was derived using a supervised image classification that utilizes training and testing data digitized from the high resolution Quickbird imagery. Image classification was performed using a decision tree classification algorithm in Idrisi Andes with a combination of multi-temporal spectral variables and elevation derived ancillary variables to help discriminate agriculture from natural vegetation (Clark Labs, 2008) (Figure 9). The spectral signatures of wheat and barley are nearly identical at the spatial and radiometric resolutions of AWiFS, and both crop types are typically planted and harvested at the same time (Currit, 2005). Therefore, the historical areal proportion of wheat to barley was derived from a seven year provincial-level agricultural database acquired from the Iraq Ministry of Agriculture and applied to the final winter grains crop classification. Research has shown that crop yield has a strong positive correlation with NDVI (Tennakoon et al., 1992; Quarmby et al., 1993; Prasad et al., 2006). An autoregressive time-series model developed in S-PLUS used MODIS NDVI time-series data and known provincial-level crop statistics to derive yield estimates. In addition, Geospatial Data Analysis Corporation (GDA Corp.) developed a semi-automated web-based yield forecaster that extracts MODIS NDVI data from the University of Maryland NDVI time-series database (http://pekko.geog.umd.edu/usda/ndvi/), and models yield forecasts based upon historical crop data. The yield forecast derived from this project showed high agreeability with the GDA Corp. automated process. http://www.gdacorp.com/

RESULTS

Interpretation of high resolution Quickbird imagery between the months of October and December 2007 revealed normal field preparation activities in the form of tillage and burning (Figure 10). However, start of season agro-meteorological data showed that temperatures in the rainfed crop region of central and northern Iraq were +5 degrees Fahrenheit above normal and that precipitation was at a 10-year record low∗; start of season conditions were determined to be unfavorable for strong crop establishment early into the crop season (Figure 11). NDVI change analysis using moderate resolution AWiFS IRS-P6 and archived Landsat ETM+ determined poor crop progression during the month of January 2008 compared to previous years. By early-February, coarser resolution MODIS NDVI time series data revealed severe declines in crop abundance compared to the 9-year average (Figure 12); Quickbird imagery confirmed near total crop failures for many of the study AOIs, especially the most important winter grains producing province of Ninawa (Figures 13 & 14). These results were extrapolated to the AWiFS IRS-P6 spatial extent to provide provincial crop assessments. An early drought warning was subsequently given to U.S. and Iraq government officials by the month of February 2008, nearly 3 months before the scheduled crop harvest, and officially confirmed by the Iraq Ministry of Water Resources during the month of April 2008. Provincial-level crop statistics were derived using the abovementioned techniques and a 42.9% decrease in average wheat area and a 58.2% decrease in average barley area were forecasted (Tables 1 & 2). Official Iraq statistics were released in the month of July and results confirmed a 92% agreement with the forecasted numbers derived from remotely sensed data.

*The last severe drought in Iraq was during the 1999/00 crop season.

Page 4: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

DISCUSSION AND CONCLUSIONS

Application of the “convergence of evidence” methodology to assess crop health and condition has been extensively used by the Foreign Agricultural Service for nearly three decades. Advances in data sources, management, access, and distribution make timely and accurate agricultural monitoring truly operational. As a result of this project early warning of drought conditions in Iraq and the possible negative effects on crop yields, provided government and security forces with the information crucial to determining the needs of the population and ensuring timely food relief before a crisis emerged. Furthermore, crop statistics derived from remotely sensed data sources provided an accurate forecast in addition to a quantitative measurement of crop production shortages for the next season.

The goal of this project was to develop new remote sensing processes that can be applied to similar environments in the region. This project continues to be funded by the National Geospatial Intelligence Agency and Phase II was implemented at the beginning of the 2008/09 winter grains season. The regional extent of the Iraq project has expanded to include the countries of Syria, Iran, and Afghanistan.∗ These countries share similar agricultural schemes with large proportions of rainfed cropland. The same methodology will be applied to crop analysis with improvements in process automation. Phase II project results will be available by June 2009.

REFERENCES Boatwright, G.O. and V.S. Whitefield, 1986. Early warning and crop condition assessment research, IEEE

Transactions on Geosciences and Remote Sensing, GE-24(1). Clark Labs, Clark University 950 Main Street, Worcester MA 01610-1477 USA. URL: http://www.clarklabs.org Colwell, R.N., 1983. Manual of Remote Sensing, II, Interpretation and Applications, Falls Church, VA: American

Society of Photogrammetry. Currit, N., 2005. Development of a remotely sensed, historical land-cover change database for rural Chihuahua,

Mexico, International Journal of Applied Earth Observation and Geoinformation, 7(3):232-247. Digital Globe, 2008. Quickbird Spacecraft. Longmont, CO, USA. URL:

http://www.digitalglobe.com/index.php/85/Quickbird Indian Space Research Organization (ISRO), 2008. IRS-P6 (Resourcesat-1). Indian Department of Space.

Bangalore, India. URL: http://www.isro.org/pslvc5/index.html Prasad, A.K., L. Chai, R.P. Singh, and M. Kafatos, 2006. Crop yield estimation model for Iowa using remote

sensing and surface parameters, Internationational Journal of Applied Earth Observation and Geoinformation, 8: 26-33.

Quarmby, N.A., M. Milnes, T.L. Hindle, and N. Silleos, 1993. The use of multitemporal NDVI measurements from AVHRR data for crop yield estimation and prediction, International Journal of Remote Sensing, 14: 199-210.

Tennakoon S.B., V.V.N. Murthy and A. Euimoh, 1992. Estimation of cropped area and grain yield of rice using remote data, International Journal of Remote Sensing, 13: 427-439.

Tetrault, R.,2006. Access and availability of Resourcesat-1 AWiFS Data for agriculture, USDA FAS Agricultural Applications Seminar, Fairfax, VA.

Tingley, W., 1988. Crop condition data retrieval and evaluation (CADRE) DBMS Dictionary, Lockheed Engineering and Sciences Company, Inc., Unpublished.

∗ Drought conditions affected the entire Middle East.

Page 5: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 1. International Production Assessment Division (IPAD) Regions of Responsibility.

Page 6: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 2. Iraq pilot study area representing the predominant agroecological zones: rainfed and irrigated.

Page 7: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 3. Five year average crop production for major winter grains: provincial-level crop statistics.

Page 8: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figures 4 & 5. Iraq crop calendar and sensor specific coverage of northern Iraq cropland.

Page 9: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 6. AWiFS IRS P-6 surface reflectance mosaic for the month of November 2007.

Page 10: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 7. AWiFS IRS P-6 NDVI mosaic for the month of March 2008.

Page 11: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 8. Hierarchical agricultural monitoring that utilizes multiple spatial resolution data: MODIS, AWiFS, and Quickbird.

Page 12: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 9. Decision tree classification results using Quickbird-derived training sites and peak crop imagery.

Page 13: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 10. Quickbird imagery revealing winter grains field preparation in the form of tillage and burning.

Page 14: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 11. Start of season agro-meteorological conditions. Adequate precipitation and temperature are crucial for the strong establishment of rainfed crops.

Page 15: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

MODIS NDVI Time-Series Data

Figure 12. MODIS time-series data compared to 5-year average and map displaying departure from 9-year average NDVI.

2007/08

2006/07

2005/06

2004/05

2003/04

Average

Legend

Page 16: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 13. AWiFS maximum value NDVI composite during peak season revealed little to no green-up in major winter grains producing provinces.

Page 17: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Figure 14. High resolution false-color infrared Quickbird imagery confirming lack of crop emergence during the peak season (i.e. total crop failure).

Page 18: DATA FUSION/INTEGRATION OF HIGH AND MEDIUM … · 2013-12-07 · ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009 DATA FUSION/INTEGRATION OF HIGH AND MEDIUM RESOLUTION

ASPRS 2009 Annual Conference Baltimore, Maryland - March 9-13, 2009

Table 1: Wheat area forecast for the 2007/08 crop season (USDA Market Year 2008/09).

Table 2: Barley area forecast for the 2007/08 crop season (USDA Market Year 2008/09).