EcoSat FAQs

EOMAP algorithms were originally calibrated for Digital Globe WorldView 2, 3, 4 Airbus Pléiades, and European Space Agency (ESA) Sentinel 2 satellite sensors; however, algorithms and pricing can be adapted for any publicly or commercially available sensor. Ask us about what sensor might be the best fit for your project.

EcoSat uses tested algorithms to define discrete areas and boundaries of different vegetation communities. These polygon “objects” are automatically given unique classification numbers. The user has control over what to name the objects or whether to delete or lump classifications and reprocess the image and data with the new classifications.

Indeed, satellites can’t “see” through clouds and cloud cover (and haze) significantly impacts the availability of quality images in image archives. However, if customers choose to “task” a satellite during a 60 day window, then the upcharge ensures a guaranteed level of quality. Our satellite provider’s use sophisticated atmospheric forecasting tools and the satellite sensor angle is adjusted to capture the best possible image in the best possible weather conditions. Once the raw image is acquired EOMAP uses atmospheric correction filters to further sharpen the image and ready it for classification.

Although it’s possible to get sub-meter resolution with satellite imagery, 2-m is the smallest resolution that EcoSat processes to balance cost, file size, and image quality. Sentinel 10-m archive images could be a low-cost alternative for characterizing vegetation over very large areas while still retaining relatively high resolution.

Depending on the sensor, there could be four to thirteen spectral bands available to analyze. Different plant species may have different spectral “signatures” that can be differentiated and typed across the different bands.

EcoSat is designed to automate the processing of imagery at scales greater than 25 km² to discern large-scale spatial patterns and trends in aquatic and wetland vegetation (e.g., system wide vegetation mapping). Hyperspectral aerial imagery acquired by drones or fixed-wing aircraft may be more suitable for small-scale aquatic or wetland vegetation restoration projects.

Yes, customers will have access to the true-color satellite imagery for download and internal use. Use and distribution of raw imagery is subject to copyright and terms provided here and by the satellite image provider.

If the image is not suitable, we will not use it. For supervised classifications where we have a good library of in situ samples, we aim for 80% accuracy. Of course, accuracy depends on local conditions and the quality/detail of the image (e.g., higher accuracy can be achieved with 8-band imagery vs comparable 4-band imagery). As spectral libraries grow with repeated surveys as well as advances in imagery sensor technology, we anticipate continued gains in accuracy and precision of vegetation classifications.

Minimum mapping unit or MMU is the specific size of the smallest feature in the map. This can be controlled by the customer during the initial order request. The minimum MMU is the resolution of the satellite sensor (e.g., you cannot get finer detail than 2-m for Airbus Pléiades imagery). However, a 100 km² image packed with 2-m polygon objects might be too “busy” or “noisy” to clearly see patterns in vegetation communities. Adjusting the MMU upward will dissolve small vegetation patches within larger beds, thereby making the major vegetation community more distinct. We will consult with you about your local management objectives and suggest an optimal MMU.

Time of year can have a significant effect on vegetation classifications. Accurate classification is most difficult during late and early times of the year when brown decayed vegetation is mixed with green vegetation. Investigators should target times of the year for acquisition when biomass is highest and color differentiation between species is greatest. Typically, this is during summer.

EcoSat automatically calculates the area of each vegetation polygon object in the survey area using custom Microsoft SQL Geography scripts and sums the area across each classification (Satellite Vegetation). If a waterbody polygon stored in BioBase’s database is intersected by the survey area, A SQL script (STintersects) automatically calculates the overlap of the processed polygon objects and the waterbody boundary and sums the areas of all classifications. This process was designed specifically for complicated spatial datasets with 10’s of thousands of polygons within polygons.

BioBase does not share our administrative waterbody boundaries and they are subject to change depending on available imagery, water level, connections to other waterbodies, or changes in administrative boundaries. If they wish, users can export the processed polygon shapefile and intersect it with their own standard shapefile to regenerate area summaries.