A SERVIR tool for analyzing changes in agriculture practices
Disclaimer: Our NDVI data provider has stopped producing a 5 day product. We are working to remedy the situation.
Agriculture interventions such as irrigation, improved fertilization, and advanced cultivars have the potential to increase food security and ensure climate resilience. However, to broaden support for such activities, environmental managers must be able to assess their impact. Field data are often difficult to obtain and decisions are made with limited information. Satellite products can provide relevant information at field- and village-wide scales to inform assessments of interventions.
Vegetation health is a good indicator of crop vigor, and several studies have connected changes in satellite-derived vegetation indices to crop yields and the length of growing season. By analyzing the vegetation index data along with a rainfall dataset, areas with and without intervention can be compared and analyzed to assess the impacts of the intervention strategy and identify similar areas for potential future intervention. Towards that end, SERVIR has developed a web-based product called AgriSERV that can assist in this procedure.
AgriSERV is a user-friendly, web-based tool that provides decision makers access to satellite- and model-derived products that can aid in assessing the impact of agriculture interventions. Two widely used datasets -- eMODIS Normalized Difference Vegetation Index (NDVI) and Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS) rainfall data -- provide the basis for the analysis. (For more information on these datasets, refer to the relevant sections below.) This tool allows users to draw (delineate) two areas of interest: one control area with no intervention and another area that has experienced intervention. A time period can be specified ranging back to the year 2000. Then, an on-demand comparative analysis can be performed. The user is presented with side-by-side charts and summary data that highlight the differences of the two areas in terms of vegetation health, derived growing season lengths, and rainfall. The analysis is based on an area weighted average of the gridded NDVI and rainfall data. The users can download the summary data table as well as the full dataset for the period specified.
The US Geological Survey Earth Resources Observation and Science EROS (USGS EROS) Center distributes a collection of satellite-derived vegetation products generated from the Moderate Resolution Imaging Spectroradiometer (MODIS). These products, known as "eMODIS," are used by operational land monitoring applications requiring near-real time NDVI data for comparison against historical records. Real-time and historical NDVI products are composited in 10-day intervals every 5 days on a geographic mapping grid.
eMODIS 10-day maximum-value composite NDVI images at 250m spatial resolution are used to monitor vegetation condition. NDVI is a measure of the density of chlorophyll contained in vegetative cover and is defined as (NIR - RED) / (NIR + RED), where NIR is the near-infrared reflectance and RED is the visible-red reflectance. This vegetation product is calculated from MODIS L1B Terra surface reflectances, corrected for molecular scattering, ozone absorption, and aerosols using MODIS Science Team algorithms.
The NDVI and NDVI anomaly maps are the products of a temporally smoothed 250m NDVI dataset. A time series smoothing technique developed by Swets et al. (1999) was used to smooth NDVI composites for the years 2001 to 2010. The technique uses a weighted least squares linear regression approach to correct observations that are of poor quality due to clouds or other atmospheric contamination This smoothed time series was used to derive a 10-year mean NDVI on a pixel-by-pixel basis for each of 72 composite periods per year. As current-year composites become available, they are added to the time series and smoothed, resulting in a smoothed composite comparable to the historical mean for a given 10-day period.
While temporal smoothing can be effective at improving time series data, it can be problematic to implement in real time for areas of persistent cloud cover. Therefore, SERVIR has implemented a process that steps back in the time series and replaces data after six composite periods, allowing cloud-free observations to have been obtained. Prior to the final corrected data being made available, interim graphics for all products are masked with cloud flags from the original input data. More information about USGS eMODIS can be found in the USGS publication (Jenkerson et al., 2010) or at FEWS NET eMODIS product page at https://earlywarning.usgs.gov/fews/product/115#documentation
Since 1999, U.S. Geological Survey (USGS) and Climate Hazards Group (CHG) scientists, supported by funding from the U.S. Agency for International Development (USAID), the National Aeronautics and Space Administration (NASA), and the National Oceanic and Atmospheric Administration (NOAA), have been developing techniques for producing historical rainfall maps, especially where surface data is sparse. This effort has led to CHIRPS, a 30+ year quasi-global rainfall dataset. Spanning 50°S-50°N (and all longitudes), starting in 1982 to near-present, CHIRPS incorporates 0.05° (~5 km) resolution satellite imagery with in-situ station data to create gridded rainfall time series for trend analysis and seasonal drought monitoring. As of February 12, 2015, version 2.0 of CHIRPS is complete and available to the public free of charge.
Estimating rainfall variations in space and time is an important aspect of drought early warning and environmental monitoring. An evolving dryer-than-normal season must be placed in historical context so that the severity of rainfall deficits can be quickly evaluated. However, estimates derived from satellite data provide areal spatial averages that suffer from biases due to complex terrain, and therefore often underestimate the intensity of extreme precipitation events. Conversely, precipitation grids produced from station data may be inadequate in more rural regions where there are fewer rain gauge stations. CHIRPS was created in collaboration with scientists at the USGS EROS Center in order to deliver reliable, up to date, and more complete datasets for a number of early warning objectives (such as trend analysis and seasonal drought monitoring).
Early research focused on combining models of terrain-induced precipitation enhancement with interpolated station data. More recently, new resources of satellite observations such as gridded satellite-based precipitation estimates from NASA and NOAA have been leveraged to build high-resolution (0.05°) gridded precipitation climatologies, which, when applied to satellite-based precipitation fields, can remove systematic bias. This was a key technique in the production of the CHIRPS dataset. The creation of CHIRPS has supported drought-monitoring efforts by FEWS NET. More information regarding CHIRPS can be found in the USGS publication (Funk et al., 2014), at the Funk et al. (2015), or at http://chg-wiki.geog.ucsb.edu/wiki/CHIRPS_FAQ.
Funk C., Peterson P., Landsfeld M., Pedreros D., Verdin J., Shukla S., Husak G., Rowland J., Hoell A. and Michaelsen J. (2015) The climate hazards group infrared precipitation with stations - a new environmental record for monitoring extremes, Scientific Data, In Press.
Funk, C. , Peterson P. , Landsfeld M. , Pedreros D. , Verdin J. , Rowland, J. , Romero B. , Husak, G. , Michaelsen, J. , Verdin A. (2014) A Quasi-global Precipitation Time Series for Drought Monitoring. USGS Data Series 832. https://pubs.usgs.gov/ds/832/pdf/ds832.pdf
Jenkerson, C.B., Maiersperger, Thomas, Schmidt, Gail, 2010, eMODIS: A user-friendly data source: U.S. Geological Survey Open-File Report 2010–1055, 10 p.
Swets, D. L., Reed, B. C., Rowland, J. R., & Marko, S. E. (1999). A weighted least-squares approach to temporal smoothing of NDVI. 1999 ASPRS Annual Conference, From Image to Information, Portland, Oregon, May 17 – 21, 1999, Proceedings: Bethesda, Maryland, American Society for Photogrammetry and Remote Sensing.