Skip to main content
U.S. flag

An official website of the United States government

AppEEARS Data Extracting Tool Drawing Rave Reviews

More often than not, researchers mining remotely sensed data would prefer not to have to download and process entire data files to investigate their scientific questions.

AppEARS data extraction tool product descriptions
The Land Processes Distributed Active Archive Center's powerful data extraction tool for Earth observation datasets, called the Application for Extracting and Exploring Analysis Ready Data Samples (AppEEARS), is drawing rave reviews. (Public domain.)

Now thanks to a data extractor called the Application for Extracting and Exploring Analysis Ready Samples (AppEEARS), which was developed by NASA's Land Processes Distributed Active Archive Center (LP DAAC) at EROS, they don’t have to.

An intuitive and powerful extraction application for Earth Observation datasets, AppEEARS allows users to study single geographic points across the planet, or go larger and sample expansive areas. They can use specified bands/layers and time ranges from scores of geospatial datasets. They can explore data values, trends, and relationships within the AppEEARS interface using several interactive charts and graphs.

By subsetting specific geographic areas, and by setting variable and temporal constraints, users are discovering that they can reduce the amount of data required to download by up to 10,000 percent in some cases—a significant savings in time and computational power.

Not surprisingly, the early returns from those users are quite enthusiastic.

AppEEARS Users Offer Good Reviews

Dr. Jeffery Thompson, a research scientist for Earth Lab at the University of Colorado, calls AppEEARS a valuable tool in the exploratory research he has done. It has benefitted him on a number of occasions in 2017, he said.

In one case, Thompson and others submitted a proposal to NASA that involved an effort to link ground observations and temperatures to remotely sensed land surface temperature as part of exploring the state of mountain permafrost. In the other instance, he and others did an analysis of land surface temperatures in Greenland to explore whether or not thawing permafrost could be linked to a landslide there in 2017.

 “It has saved me a lot of time,” Thompson said of AppEEARS and his work on those efforts. “It’s taken all the headache and heartache out of the things that used to take large cycles.”

The idea for AppEEARS came out of the LP DAAC’s User Working Group, a consortium of users from universities, federal agencies, and other stakeholders who were interested in addressing a number of challenges in their work. They wanted a way to sample data based on specified administrative or landscape units—say a watershed, for example, or maybe a region—some of which may come from different coordinate reference systems and even different formats.

Once they had the point or area identified, they wanted to look at variables of interest within those points, or areas that maybe were only available through different data products, and then bundle them all together.

“So that challenge was really about … dataset interoperability,” said Tom Maiersperger, the AppEEARS principal investigator at EROS. “It really became about bringing these things together in a harmonized way, and particularly bringing those things together for that area you’re interested in.”

AppEEARS Accesses Variety of Data

As AppEEARS has matured to be able to sample points and areas, and to bring some harmonization to the different datasets, users were interested in bringing as much different data to the mix as possible. So from MODIS, Web-Enabled Landsat Data (WELD), and elevation datasets only available from the LP DAAC, AppEEARS began expanding out to include data products from other DAACs in the same Earth Observing System Data and Information System (EOSDIS). Today there are more than 100 datasets available to AppEEARS users.

 “We’re not bringing these datasets into our system and managing them,” Maiersperger said. The other DAACs “continue to manage those datasets within their organizations, but they present those datasets in a way that AppEEARS can see them, locate them in time and space, and essentially pull the pieces and parts of those data into the service layer, and then work on them and stage them for retrieval by the user.”

The AppEEARS team has added MODIS snow data products from the National Snow and Ice Data Center (NSIDC) DAAC, and is working to add a selection of datasets from NASA’s Soil Moisture Active Passive (SMAP) mission, which are also distributed by the NSIDC DAAC. There are population data products from the Socioeconomic Data and Applications Center (SEDAC). And meteorological products like precipitation and temperature measurements from Daymet are in the works from the Oak Ridge National Lab. There’s even work being done to bring Landsat Analysis Ready Data (ARD) products to AppEEARS.

“What we’ve shown is that in the case of the Landsat ARD surface reflectance products, we can indeed … get them arrayed and exposed in a way that we can pull those pieces and parts into the AppEEARS processing chains and put out stuff on the other side that looks as we would expect,” Maiersperger said.

The reality is, users wanting to produce a global Landsat product that leverages all the best data in the Landsat archive are not going to use AppEEARS as a large-scale processing system, Maiersperger said. People who try to compare what AppEEARS does to, say, a Google Earth Engine or a Data Cube application or even a data or information warehouse, are missing what the tool is all about.

AppEEARS Samples Points, Areas

AppEEARS really is just a sampler, Maiersperger reiterated. With the Point Sampler that went out with the initial AppEEARS release in March 2016, and the AppEEARS Version 2.0 with the Area Sampler and additional datasets that came out in September 2017, users will find that using the data sample extraction tool can make data access and processing much easier. Additionally, the at-archive data reduction will be more robust, the data transformation more efficient, and the exploratory visuals more informative.

The work that Thompson at Earth Lab has done in Greenland both before AppEEARS became available and after its development has given him an appreciation of its utility.

Before AppEEARS, Thompson and another colleague had been doing work in areas of Greenland where snow was staying on the ground for shorter periods of time, and thus vegetation had more time to grow. In looking at the question of whether some of that vegetation was drying out, they turned to a number of data products to help come up with an answer, including a MODIS daily snow cover product, MODIS surface temperature products, MODIS NDVI, and more.

As it turned out, the first month of that effort was probably consumed with simply downloading the data, Thompson said. He had written a code for his doctoral dissertation that stitched two MODIS tiles together, but in trying to extend that to include the whole of Greenland, “a lot more time was consumed just getting my codes working, and getting the tiles mosaicked, subsetted, and stitched together,” he said.

More recently, he and others did an analysis of land surface temperatures in Greenland to explore whether or not thawing permafrost may have contributed to a landslide that caused a tsunami there in 2017. “For that analysis, I did use AppEEARS, which really accelerated the analysis,” Thompson said. “In that case, I was able to request the data, download it, and process it within a very short time-frame than would have been possible had I tried to adapt my old workflow.”

Going forward, Maiersperger said his AppEEARS team is working to answer the user community’s desire for an Application Programming Interface—known as an API—that makes it easier for users to directly use AppEEARS’ services. They also are refactoring codes and making other changes to someday enable the move of AppEEARS capabilities to the Cloud.

And of course, as best they can, they’re always interested in adding more datasets.

“We like where we’re going,” Maiersperger said. “In addition to more datasets, there are some small improvements we need to make to the User Interface, and we need to do some bug squashing. You’ve always got to do a little bit of that, but we’re pleased with what we’ve got. We think we’ve got a very good product.”