Does it matter whether public-oriented decision processes and outcomes are transparent, accountable and fully informed? Participatory decision making is as old as democracy.From national elections to local land-use planning, participatory decision making exemplifies the democratic maxim that those affected by a decision outcome should participate directly in decision processes.

For years, a GIS has been called a decision support system (Cowen 1988).When we connect GIS with participatory contexts we get "participatory GIS," (PGIS) (Harris et al.1995, Jankowski and Nyerges 2001) or a "public participation GIS" (PPGIS) (Nyerges, Barndt, Brooks 1997), or community integrated GIS (Craig, Harris, and Weiner 2001). The term public is defined by the American Heritage Dictionary Fourth Edition as "the community or the people as a whole or alternatively a group of people sharing a common interest as in the reading public.The term participation is defined as the act of taking part or sharing in something." Given that the term public is an adjective qualifying participation in PPGIS, the term PGIS is actually the broader, more inclusive, of the two phrases - although some researchers prefer to restrict use of "PGIS" to participatory development within developing countries. In the late 1990's, the phrase "community integrated GIS" became preferred to refocus research effort on community (Craig, Harris and Weiner 2001).Whether we use "participatory" or "public participation" or "community integrated" does not matter, as public is commonly constituted of diverse groups with varying interests.

Regardless of the label, individuals as part of the public and groups within the public are often marginalized in public decision processes. Groups/communities traditionally marginalized, i.e., under-represented in decision processes at macro scales, are now organizing with the help of action research projects.For example, within the City of Milwaukee, in a project called Making Connections, residents helped with GIS maps to form a story about community development potential.In another example, drawing on local knowledge of indigenous peoples, participatory 3D modelling helps visualize landscapes for protected area management in Southeast Asia.In a third example, local knowledge from Native American tribes is sensitive to tribal cultural history.Unfortunately, GIS data recording of such knowledge could be exploited for development purposes and/or to protect culturally sensitive sites.In the studies above, as well as in most PPGIS activities, participation focuses on smaller groups rather than larger groups.

When examining an ability to give the public a voice in democracy, marginalized voice is a fairly pervasive problem.Practically speaking, the general public is constituted of many diverse groups - even if we consider the public as a whole.The general public is actually a marginalized group when it comes to participation processes, as there is no single, directed voice in the public.

Despite many federal, state and local laws that require public participation, research about local governance indicates that large-group participation in publicly oriented decisions commonly involves little meaningful participation.Meaningful participation can be defined in terms of access to voice (a deliberative process) and competence of knowledge(s) (an analytic process) that fosters shared understanding about values, interests and concerns that underlie the recommendations/choices to be offered/made by those with a stake in the decision (National Research Council 1996). Meaningful participation is a hallmark of a healthy democracy, particularly deliberative democracy in contrast to representative (make a vote) democracy.

Deliberative democracy involves empowerment wherein a reasoned discussion among people promotes shared understanding on a topic followed by consensus building.Although interest in deliberative democracy has existed for over 100 years (Gastil and Levine 2005), research and practice has blossomed since the late 1980's.Over the past decade, hundreds of deliberative democracy events of varying sizes have occurred across the world.A synthesis of case studies appears in a Deliberative Democracy Handbook (Gastil and Levine 2005).Several of the chapters deal with location-based issues, and thus GIS could be useful.However, no chapters actually refer to GIS, a seeming disconnect and latent opportunity.Of special note, some deliberative events have involved thousands of people at once.In 2002, AmericaSpeaks, a not-for-profit organization located in Washington D.C., organized a 21st Century Town Meeting at the Javitt's Center in New York City that brought together more than 3,000 people to discuss redevelopment plans for the World Trade Tower site after September 11, 2001.Unfortunately, such activity comes at great expense, requiring hundreds of facilitators at multiple levels of responsibility.

Research about analytic-deliberative decision processes has shown that meaningful public participation is possible and decision outcomes are improved (National Research Council 1996).The analytic component provides technical information that ensures broad-based, competent perspectives.GIS has provided technical information in such processes as maps can represent changes in landscapes.The deliberative component provides an opportunity to give voice to choices about values, alternatives and recommendations. Unfortunately, such public participation has been expensive and time consuming, and involved small to medium-sized groups (10-15 people). Working through analytic-deliberative participation in small to medium-sized groups in face-to-face settings is a start, but scaling analytic-deliberative participation to large groups is a challenge - as scale matters.

In addition, whether groups are better supported in face-to-face settings or in asynchronous settings is still an open research question.It is often thought that face-to-face participatory settings are superior to asynchronous participatory settings.It only seems reasonable.However, Dowling and St.Louis (2000) have shown that an asynchronous nominal group process was more effective than a face-to-face nominal group process, at least in a small group setting - a challenge to anecdotal feelings about face-to-face participatory processes.

Based on the following three observations: (1) public participation is mandated by many federal, state and local laws encouraging core democratic process; (2) the Internet is growing in popularity and access is getting better even for under-represented groups as reported in several studies; and (3) asynchronous, structured participation methods have been shown to be at least as good and in some cases superior to face-to-face participation; perhaps an Internet platform combining GIS (i.e., data management, spatial analysis and geovisualization) technologies, decision modeling technology, and communications technology into a geospatial portal to support an analytic-deliberative process might be one way to foster meaningful participation in large groups as well as hold down the cost to all who wish to participate.This rationale is the basis of the US National Science Foundation-funded research activity called the Participatory GIS for Transportation (PGIST) Project.The principal research question for the project is: What Internet platform designs and capabilities, particularly including GIS technology, can improve public participation in analytic-deliberative transportation decision making within large groups?

Addressing that question involves intensive interdisciplinary work. From experience on several large interdisciplinary projects over the past decade, getting researchers to work with one another has been equally as challenging as getting stakeholders to work together.The commonality of that challenge and the one for analytic-deliberative public participation involves resolving mismatch of language meanings, e.g., the meaning of transportation-based valued-concerns. Computer-supported resolution of mismatched meanings among many people is at the core of the grand challenge.Certainly, this essay is not the first to recognize this challenge.Other literature, e.g., dealing with boundary objects in science and technology studies, recognize a similar challenge.The difference here, however, is in recognizing that scale really matters, and is particularly constraining when working in a computer-mediated setting.For those who see the World Wide Web as promoting anytime, anywhere and anytime work, it just isn't so as yet in connection with analytic-deliberative work at large, participatory scales - what just might be considered a grand challenge for PPGIS work.

Cowen, D.1988.GIS versus CAD versus DBMS: what are the differences? Photogrammetric Engineering and Remote Sensing, 54, 11, 1551-1555.

Craig, Will, Trevor Harris and Daniel Weiner (Editors) (2001) Community Participation and Geographic Information Systems. Taylor and Francis.

Dowling, K.L.R.D.St.Louis 2000.Asynchronous implementation of the nominal group technique: is it effective? Decision Support Systems 29:229:248.

Gastil, J and P.Levine 2005.The Deliberative Democracy Handbook.San Francisco: Jossey-Bass.

Harris, Trevor M., Weiner, Daniel, Warner, Timothy A., and Levin, Richard.1995.Pursuing social goals through participatory geographic information systems.Redressing South Africa's historical political ecology.In Pickles, J., ed., Ground Truth: The Social Implications of Geographic Information Systems. New York: Guilford Press.196-222.

Jankowski, P.and Nyerges, T.2001.Geographic Information Systems for Group Decision Making: Toward a Participatory, Geographic Information Science, London, Taylor & Francis.

National Research Council 1996.Understanding Risk: Informing Decisions in a Democratic Society, Washington, D.C., National Academy Press.

Nyerges, T, M.Barndt, and K.Brooks, 1997. Public Participation Geographic Information Systems, Proceedings, Auto-Carto 13, Seattle, WA, American Congress on Surveying and Mapping, Bethesda, MD, pp.224-233.

Articles

Scaling-Up as a Grand Challenge for Public Participation GIS

Ed. note: This article originally appeared on the National Snow and Ice Data Center's Highlights website.

NSIDC and NASA data scientists proved the use of 21st-century techniques to revive data from 1960s satellites.

Scientists today who study polar sea ice conditions rely on satellite records reaching back to 1979. But soon, data scientists hope to extend the look back by another decade or more. Researchers at NSIDC and NASA have shown that the oldest Earth observing satellite data can be made to yield new information, adding significantly to the view of Earth's climate history.

When NASA launched the first Nimbus satellite in the 1960s, they also launched an era of Earth observations from space. While the early Nimbus satellites provided meteorological and other observations, methods did not yet exist to detect features such as the margins of the sea ice cover in the Arctic and Antarctic. Even if they had, the limits of computer processing in those days would have made quantitative analysis unfeasible.

These early satellite data still reside in NASA archives on archaic, two-inch tape media. When NSIDC scientist Walt Meier and project manager Dave Gallaher learned that NASA researchers had retrieved 1960s images of Earth from the Lunar Orbiter, they wondered if early NASA satellite data could also yield information about sea ice conditions before 1979. They saw a disappearing window of opportunity to recover these data. Only one tape drive remained in the world that could read the Ampex two-inch media. Plus, the original Nimbus researchers were now in their late 70s and 80s, and contact with them would be critical to answering some of the necessary instrumentation questions.

On September 23, 1966, Nimbus II acquired the temperature data shown in this image of Asia, overlaid on Google Earth. The warm waters of the Persian Gulf are clearly seen in contrast with the cold temperatures of the Himalayas. Blues and greens indicate cooler temperatures, and yellows and reds indicate warmer temperatures.

The Nimbus satellites carried a High Resolution Infrared Radiometer (HRIR), with a resolution of 8 kilometers, a medium-resolution infrared radiometer (MRIR), and an Advanced Vidcon Camera System (AVCS). Although Nimbus collected data for only short periods, Meier suspected that these early Nimbus satellites likely captured the annual Arctic sea ice minimum, occurring each September.

Starting with the methods developed for the Lunar Orbiter Image Recovery Project (LOIRP) at NASA Ames Research Park, a team at NSIDC worked with Dennis Wingo at LOIRP to search NASA archives for the original Nimbus tapes containing raw images and calibrations. Their first goal was to read and reprocess the data at a higher resolution, removing errors resulting from the limits of the original processing.

These tasks proved more challenging than expected, due to truncated data, missing algorithms, and other issues. But the result was a global image of the Arctic from Nimbus II, captured on September 23, 1966, in higher resolution than ever seen before from this type of data. This date falls around the time that Arctic sea ice would have reached its end-of-season minimum extent. The image demonstrates the possibility of reprocessing the entire available time series, supporting new scientific study of past conditions on Earth.

This proof-of-concept research was supported by an Innovative Research Program seed grant from the Cooperative Institute for Research in Environmental Science (CIRES) at the University of Colorado at Boulder, the home of NSIDC.The team now plans to apply for further research funding to expand the study. More work is needed to reprocess 2,600 half-orbit records, develop a de-jitter algorithm, and then complete a painstaking process of image comparison to separate clouds from ice.

Processing of more data could allow Meier and other scientists to determine sea ice extent during the Nimbus I through III campaigns (1964, 1966, and 1969). The team also hopes to extend their methods to later instruments, and characterize both minimum Arctic and maximum Antarctic sea ice extent for 1964 to 1978. If successful, the sea ice climatologies would gain up to fourteen years of data - an increase of 50 percent over their temporal span today.

Techno-Archaeology Rescues Climate Data from Early Satellites: Archived but not Forgotten

You may have noticed that “copyright,” “public domain,” and “1923” have been in the news lately. At midnight on Dec. 31, 2018, certain intellectual property published in 1923 that had been held under copyright underwent a legal transformation and is now part of the public domain. For a while this was an annual occurrence within the copyright lifecycle, since earlier instances of copyright laws had provided protection for authors, artists, and other creators for a fixed number of years (50 years after the author’s death, for example). This cyclical advancing pattern of release was interrupted in 1998 when the Copyright Term Extension Act became law, essentially putting a freeze on the otherwise annual release of things by automatically extending the number of years of protection. Mickey Mouse, Sonny Bono, and the entertainment industry had a lot to do with this, but the impact was necessarily widespread across all materials.

Specifics of copyright protection are complex and detailed, and necessarily so, as the range of original creations worldwide is equally rich. The laws allow individuals and corporations to know that their precious efforts are recognized as such. At the same time, some of our most beloved music, movies, artworks, and the like were directly inspired by previous publications. Imitation is the sincerest form of flattery! Once something is part of the public domain, it can be used, reused, altered, modified, and reworked. It is available to become the raw stuff of inspiration for other creators, which is why some museums are blazing a trail into the open access realm.

Moreover, just because something exists under copyright doesn’t mean one can’t actively use it. For example, the London Underground regularly permits – through authorized licensing – alternative representations of the famous 1931 Harry Beck tube map to be published without legal sanctions. Licensing is what allows still-under-copyright music to stick in our heads in commercials. The standardized language of Creative Commons licenses has been a valuable solution for those wishing to facilitate the sharing of content for creative rather than commercial means, such as the David Rumsey collection of maps.

Meanwhile, think of Jan. 1, 2019, as having been a flush of new public domain content as the release cycle began anew. Books, movies, visual art, and music have received most of the attention, but what about maps? As a reminder, almost all maps published under the auspices of the U.S. Government fall outside of these restrictions anyway, as they are part of the public domain from their beginning. So the phenomenal maps of the Mississippi River that Harold Fisk produced in the early 1940s for the U.S. Corps of Engineers have always been available – and now in digital form – for your use as art inspiration, background images, or screen printing onto t-shirts.

Out-of-copyright (or never under copyright) historical maps exist for many locales. For example, many of the Sanborn Fire Insurance maps, a large and valuable collection for the United States that date back to the year 1866, are slowly and steadily being scanned and becoming available for download from the Library of Congress, at least up to their current year of copyright protection. However, becoming-public-domain at midnight on a certain date is not as similar to the Cinderella-midnight-pumpkin analogy as it may seem. Copyrights could be, and were, extended past these threshold dates when and if authors or publishers registered and then extended the copyright within a particular time frame. If such registration and extension is known with certainty not to have happened, then more recent items may have lost their copyright protections already…except that after 1991, the registration renewal requirement was waived. Really, it’s quite complicated.

There are many layers of politics and legalities surrounding these matters, and few cases are black and white. Prepare to do research, especially when there are no obvious clues to a map’s source and publication date. The U.S. Copyright Office offers suggestions for how to begin this type of investigation (a 12-page PDF document not yet updated with the new 1923 date). And what about new forms of spatial data and novel representations? What does it mean for LiDAR enthusiasts that a 3-dimensional scan may not qualify for protection under copyright? CyArk already shares some of its LiDAR data. Will others?

Are there specific maps published in 1923 that have just entered the public domain? National Geographic was actively producing maps during those years, and their 1923 map of the United States beautifully documents national activities from that time, such as highway expansion. Is this particular map suddenly available to be done with what we wish, by virtue of its alleged publication date? Not so fast. It turns out that this map was, indeed, first published in April 1923, but currently it isn’t known with certainty whether National Geographic previously extended its copyright. Plus, we do not know about licensing arrangements with Art.com, through which National Geographic distributes its historical maps. Can this 1923 map be purchased as a paper reproduction, scanned at a high resolution, and its digital representation then georeferenced to become a basemap within a GIS, with or without requiring attribution or further permission? Inquiring minds want to know, and lawyers will get back to us. Like I said, it’s complicated.

Copyright, Public Domain, and Maps – It’s Complicated

Summary: In this article, Luca Congedo presents a tutorial for the land cover classification of cropland. In particular, he uses Landsat imagery acquired over the State of Kansas, near the city of Ulysses, using the new version 2.3.2 of the Semi-Automatic Classification Plugin for QGIS.

In this post, I am presenting you a tutorial for the land cover classification of cropland. In particular, we are going to classify a Landsat image acquired over the US state of Kansas, near the city of Ulysses, using the new version 2.3.2 of the Semi-Automatic Classification Plugin for QGIS.
 
Before the tutorial, please watch the following video that illustrates the study area and provides very useful information for the interpretation of the image in false colors (footage courtesy of European Space Agency/ESA). Also, a brief description of the area that we are going to classify is available here.As shown in the previous video, the study area is mainly covered by cropland, and we are going to classify a Landsat 8 image acquired in April 2013 (available from the U.S. Geological Survey), using a semi-automatic approach (i.e. the entire image is classified automatically basing on several samples of pixels "ROIs" that are used as reference classes).

Before the classification process, we are going to convert the DN values to reflectance. The reflectance is defined as the ratio between reflected and incident energy on a surface (see here for more information about the conversion to reflectance of Landsat images).

The atmospheric correction of reflectance values is useful during the pre processing phase, because the electromagnetic energy measured by a satellite is affected by the atmospheric effects (e.g. absorption or scattering). Therefore, image conversion to reflectance allows for the comparison and mosaic of different satellite images (e.g. Landsat 5 and Landsat 8), and improves the classification results.

The Semi-automatic Classification Plugin provides a simple correction using the DOS1 method (a brief explanation about image conversion to reflectance and DOS1 atmospheric correction is illustrated in the yellow frame at the end of this post). It is worth mentioning that GRASS GIS has also a tool for  the calculation of the Top Of Atmosphere reflectance (see i.landsat.toar) and other DOS methods.

First, download the image from here. The image is a subset of the entire scene, including the following Landsat bands (each band is a single 16 bit raster):

The following workflow illustrates the main phases of the semi-automatic classification:


1. Conversion of raster bands from DN to reflectance

This phase is part of image pre processing, which enhance the input images by converting the DN (i.e. Digital Numbers) to the physical measure of Top Of Atmosphere reflectance (TOA). Also, we are going to apply a simple atmospheric correction using the DOS1 method (Dark Object Subtraction 1), which is an image-based technique.

As an example, the following images illustrate three charts of different spectral signatures (water, vegetation, and bare soil) calculated with the original image (DN), the TOA reflectance image and the DOS1 corrected reflectance image. As you can see, the DOS1 correction modifies the reflectance values, especially in the blue band. Notice that the band number in the following images is the numerical order of used bands, thus for example: band 1 of the chart is band 2 of Landsat 8 (i.e. blue band).

Open QGIS and start the Semi-Automatic Classification Plugin; in the main interface select the tab Pre processing > Landsat;

Select the directory that contains the Landsat bands (and also the required metafile MTL.txt), and select the output directory where converted bands are saved;

Check the option Apply DOS1 atmospheric correction, and click Perform conversion to convert Landsat bands to reflectance;

At the end of the process, converted bands are loaded in QGIS.

2. Definition of the classification inputs

Now, we create a color composite of the image. In particular, the composite RGB = 543 (that is the equivalent of RGB = 432 for Landsat 7) is useful for the interpretation of cropland in the image (as shown in the previous video by ESA), because of healthy vegetation reflects a large part of the incident light in the near-infrared wavelength.

In order to create the color composite in QGIS, we create a virtual raster. Then, we define the band set that we are going to classify. Finally, we create a training shapefile which stores the ROIs (i.e. Regions Of Interest) that define the land cover classes.

From the menu Raster select Miscellaneous > Build Virtual Raster (catalog); click the button Select... and select the bands 3, 4, and 5; select the output file (for instance rgb.vrt); check Separate (bands will be separated) and click OK;

In the plugin dock ROI creation click the button Band set beside Select an image; the tab Band set will open; click the button Select All, then Add rasters to set (order the band names in ascending order, from top to bottom using ↑ and ↓ arrows);

In order to create the training shapefile (a polygon layer having several attribute fields that will store ROIs), in the dock ROI creation click the button New shapefile, and select where to save the shapefile (for instance ROI.shp).

We need to create several ROI, considering the spectral variability of land cover classes. It is important that ROIs represent homogeneous areas of the image, therefore we are going todraw the ROIs using a region growing process (i.e. a segmentation of the image, grouping similar pixels).

The following are the land cover classes that we are going to identify in the image:

low vegetation (e.g. fields without green vegetation, or shrubland)

built-up (e.g. artificial areas, buildings and asphalt)

As you can see in the color composite, the color of cropland can vary from dark red to light red. We should create several ROIs for each of these colors.

In order to create a ROI, in the dock ROI creation click the button + beside Create a ROI and then click any pixel of the image; zoom in the map and click on a red pixel of a circular field; after a few seconds the ROI polygon will appear over the image (a semitransparent orange polygon);

Under ROI definition type a brief description of the ROI inside the field Information (e.g. crop; notice that descriptions are not used during the classification process, but are useful for distinguishing ROIs); the field ID (i.e. identifier of the class) is used as reference for the land cover classification, therefore it is important that each category has a unique ID value (ROIs sharing the same ID are evaluated as a single ROI); since we need more than one ROI per class, set the ID = 11, so that the first figure represents the class, and second one the ROI number (other crop ROIs will have values from 11 to 19; in the same way we set ID from 21 to 29 for low vegetation, and so on);

In order to save the ROI to the training shapefile click the button Save ROI to shapefile;

In the plugin main interface, select the tab ROI tools > Spectral signature, which displays the plots of selected ROIs, and select the item crop; if the checkox Plot σ is checked, then the plot will display the standard deviation of each ROI.


Repeat the above steps for every class (in order to get the desired ROI, change the values for Minimum ROI size and Range radius); after the collection of several ROIs, it is useful to visualize the ROI spectral signatures in order to avoid the collection of ROIs that are spectrally similar. Following, some examples of ROIs for the land cover classes.

4. Classification of the study area

It is useful to perform classification previews, in order to assess the quality of classification (for example a common problem is soil classified as built-up and vice versa). If the result of the preview is not good, then we need to go back to the phase of ROI creation, delete the faulty ROIs and/or create new ones. 

The classification can be performed through several algorithms; this time we are going to use the Spectral Angle Mapping classification.

In the Classification dock, under Classification preview set Size = 500 (i.e. the side of the classification preview in pixel unit); click the button + and then click on the image; after a few seconds, the classification preview will be loaded (with default colors);For a better interpretation of classes, it is useful to assign a classification color to each class.

Also, we can save a visualization style in QGIS, and set this .qml file as the default style for every classification or preview.

In the panel Layers, left click on the classification preview and select Properties > Style; select the render type Singleband pseudocolor and change the colors and labels of classes, according to the training ROIs; then, click Save style ... to save the .qml file (e.g. classification.qml);

Under Classification style click the button Select qml to select the file classification.qml; thus the next classification or preview will be loaded with this style;When we are satisfied with the result, we can perform the final classification (the main output of a classification is a raster file .tif).

click the button Perform classification and select where to save the output (e.g. classification.tif).

5. Calculation of classification accuracy

It is useful (and necessary) to assess the accuracy of land cover classification, in order to understand the reliability thereof, and to identify map errors.

Please, notice that this classification is only for demonstration purposes, because an accurate land cover classification would require ancillary data and field survey.

However, in order to assess the classification accuracy, we can compare the land cover classification to the ROIs we have created (theoretically, image pixels belonging to a certain ROI should be classified as that ROI's class).

Select the tab Post processing > Accuracy  of the plugin main interface; select the classification.tif beside Select a classification to assess  and select the ROI shapefile beside Select the reference shapefile; then click the button Calculate error matrix and the matrix will be displayed; you can save the error matrix by clicking the button Save error matrix to file; Generally, classification accuracy is considered good if it is > 80%. For this classification, we can see that the overall accuracy of the classification is about 91.4% (the error matrix is reported below). Also, it is useful to consider the error for single classes (i.e. omission and commission).

6. Calculation of classification statistics

The classes of the final classification are more than the main land cover classes we have defined at step 3 (this is because we needed to assign diverse IDs also to ROIs belonging to the same class).

In order to estimate the cropland in the area, we are going to reclassify the classification raster in fewer classes, according to the main land cover classes.

From the Processing toolbox navigate to SAGA > Grid - Tools > Reclassify Grid Values; in the tool window, under Grid selectclassification.tif; under method select [2] simple table;

Under Lookup table click the button ... to open the window Fixed Table; the lookup table has 3 columns: minimum, maximum, and new. Minimum and maximum define the reclassification range for the new value, according to the following table. 

Click "OK" and after a few seconds the reclassified classification will be loaded. Now we can perform the report of the classification, which provides the percentage and the area of land cover classes.

Select the tab Post processing > Classification report  of the plugin main interface; select Reclassified grid beside Select a classificationand click Calculate classification report; after a few seconds the report will be displayed (see the following table).

From these results, we can see that in the study area (at the moment of image acquisition) there are about 738 square kilometers of crop class (class 1, about 10% of the study area) and 6231 square kilometers of low vegetation class (class 2, about 87%). Also, we can notice that about 14 square kilometers are occupied by farms (class 4) and about 18% is built-up (class 3).

Of course, these numbers are the result of a tutorial for demonstration purposes; for a better estimation of class areas (and possibly the identification of the crop type), we need field data which can improve the creation of ROIs and help for the assessment of classification accuracy.

However, the aim of this tutorial is to show how land cover classifications can be performed simply and rapidly using open source programs, which are useful for several activities, such as land cover monitoring or precision farming.

Landsat image conversion to reflectance and DOS1 atmospheric correction

Radiance is the “flux of energy (primarily irradiant or incident energy) per solid angle leaving a unit surface area in a given direction”, “Radiance is what is measured at the sensor and is somewhat dependent on reflectance” (NASA, 2011, p. 47).

The Spectral Radiance at the sensor's aperture (Lλ ) is measured in [watts/(meter squared * ster * μm)] and for Landsat images it is given by (https://landsat.usgs.gov/Landsat8_Using_Product.php):

- ML = Band-specific multiplicative rescaling factor from Landsat metadata (RADIANCE_MULT_BAND_x, where x is the band number)

- AL = Band-specific additive rescaling factor from Landsat metadata (RADIANCE_ADD_BAND_x, where x is the band number)

- Qcal = Quantized and calibrated standard product pixel values (DN)

“For relatively clear Landsat scenes, a reduction in between-scene variability can be achieved through a normalization for solar irradiance by converting spectral radiance, as calculated above, to planetary reflectance or albedo. This combined surface and atmospheric reflectance of the Earth is computed with the following formula” (NASA, 2011, p. 119):

ρp = (π * Lλ * d^2 )/ (ESUNλ *cosθs)

- ρp = Unitless planetary reflectance, which is “the ratio of reflected versus total power energy” (NASA, 2011, p. 47)

- Lλ = Spectral radiance at the sensor's aperture (at-satellite radiance)

- d = Earth-Sun distance in astronomical units (provided with Landsat 8 metafile, and an excel file is available from

  http://landsathandbook.gsfc.nasa.gov/excel_docs/d.xls)

- ESUNλ = Mean solar exo-atmospheric irradiances

- θs = Solar zenith angle in degrees, which is equal to θs = 90° - θe where θe is the Sun elevation

It is worth pointing out that Landsat 8 images are provided with  band-specific rescaling factors that allow for the direct conversion from DN to TOA reflectance. However, the effects of the atmosphere (i.e. a disturbance on the reflectance that varies with the wavelenght) should be considered in order to measure the refletance at the ground. As described by Moran et al. (1992), the land surface reflectance (ρ) is:

ρ = [π * (Lλ - Lp) * d^2]/ [Tv * ( (ESUNλ * cosθs * Tz ) + Edown )]

- Tv is the atmospheric transmittance in the viewing direction

- Tz is the atmospheric transmittance in the illumination direction

- Edown is the downwelling diffuse irradiance

Therefore, we need several atmospheric measurements in order to calculate ρ (physically-based corrections). Alternatively, it is possible to use image-based techniques for the calculation of these parameters, without in-situ measurements during image acquisition.

The Dark Object Subtraction (DOS) is a family of image-based atmospheric corrections.

Chavez (1996) explains that “the basic assumption is that within the image some pixels are in complete shadow and their radiances received at the satellite are due to atmospheric scattering (path radiance). This assumption is combined with the fact that very few targets on the Earth's surface are absolute black, so an assumed one-percent minimum reflectance is better than zero percent”. It is worth pointing out that the accuracy of image-based techniques is generally lower than physically-based corrections, but they are very useful when no atmospheric measurements are available as they can improve the estimation of land surface reflectance.

The path radiance is given by (Sobrino, et al., 2004):

- Lmin = “radiance that corresponds to a digital count value for which the sum of all the pixels with digital counts lower or equal to this value is equal to the 0,01% of all the pixels from the image considered” (Sobrino, et al., 2004, p. 437), therefore the radiance obtained with that digital count value (DNmin)

- LDO1% = radiance of Dark Object, assumed to have a reflectance value of 0,01

The radiance of Dark Object is given by (Sobrino, et al., 2004):

LDO1% = 0,01* [(ESUNλ * cosθs * Tz ) + Edown] * Tv / (π * d^2)

Lp = ML* DNmin + AL – 0,01* [(ESUNλ * cosθs * Tz ) + Edown] * Tv / (π * d^2)

There are several DOS techniques (e.g. DOS1, DOS2, DOS3, DOS4), based on different assumption about Tv, Tz , and Edown .

The simplest technique is the DOS1, where the following assumptions are made (Moran et al., 1992):

Lp = ML* DNmin + AL – 0,01* ESUNλ * cosθs / (π * d^2)

And the resulting land surface reflectance is given by:

ρ = [π * (Lλ - Lp) * d^2]/ (ESUNλ * cosθs) 

ESUN [W/(m2 * μm)] values for Landsat sensors are provided in the following table.

* from Chander & Markham (2003)

For Landsat 8, ESUN can be calculated as (from http://grass.osgeo.org/grass65/manuals/i.landsat.toar.html):

ESUN = (π *d2) * RADIANCE_MAXIMUM / REFLECTANCE_MAXIMUM

Land Cover Classification of Cropland: a Tutorial Using the Semi-Automatic Classification Plugin for QGIS

By uniting genuine volumetric detail for 360° coverage of pit walls with a longer operating reach and a built‑in panorama view, the platform enables quicker discovery of movement, stronger trust in warnings, and sharper risk evaluations.

Industry-First GB SAR With Volumetric Precision

Through a new release from its parent group, the radar unit has introduced ArcSAR Neo under the IDS GeoRadar banner, a ground-based GB SAR platform delivering authentic volumetric detail for all-around open-pit observation and monitoring of slopes. By pushing SAR techniques further, this release supplies dependable, high‑grade radar data, strengthening risk control and enabling better choices for management and mine safety across demanding sites.

Where competing GB SAR installations typically reconstruct depth from flat radar slices, ArcSAR Neo records volumetric information directly during observation runs. By acquiring full volume natively, the sensor cuts interference at the origin and lifts the level of data clarity. With troublesome echoes and passing equipment more effectively suppressed, movement estimates become more trustworthy; crews act faster and avoid needless work stoppages.

360° Visual Context, Range, and Fast Setup

In addition to volumetric collection, the built‑in panoramic camera with HDR supplies geo‑referenced, full‑sphere (360°) visuals that are time‑aligned with the radar volume, giving operators clear context while reading the radar data. Higher resolution and a reach of as far as 6 km help with early detection of subtle movements and fast‑developing bench-scale issues. As engineered for field use, the trailer‑based configuration allows one technician to finish deployment in about a quarter of an hour.

When sunlight fades or weather lingers, photovoltaic panels can run ArcSAR Neo entirely on solar power; for continuity, a diesel supply can be selected instead.

Ids Georadar Launches Arcsar Neo

Across varied use cases, a newly designed, flexible platform fuses state-of-the-art imaging and LiDAR to raise capture efficiency by about thirty percent.

From Hexagon’s portfolio, Leica Geosystems is introducing Leica CityMapper-3. Built with adaptability in mind, the solution is aimed at speeding collection workflows and simplifying data capture for challenging regional and urban survey work.

Configurable Imaging & LiDAR for Airborne Mapping

Debuting within the visual payload is a Leica MFC250–driven suite using Sony’s IMX811 sensor, and capture throughput can rise by roughly thirty percent versus earlier units. Using a modular, multi-profile approach, crews can tune line width and ground detail to altitude and job goals, while refinements to vertical and slanted capture sharpen features from every viewpoint, enabling outputs from city-scale replicas to 2D ortho maps.

The visual payload is paired with the Leica Hyperion3+ unit, which mirrors that flexibility; pattern options are tuned to terrain and flight plans. Inside one assembly, the operation of optics and the channel for lidar is tightly aligned, yielding consistent, analysis-ready datasets in fewer sorties—a time saver for teams operating airborne sensors.

Meeting the Demand for Airborne Mapping Solutions

For teams that need flexibility, CityMapper-3 answers the call for a tool that pivots quickly across varied scenarios without trading away result quality. With one hybrid sensor, organizations can address many jobs while keeping the level of fidelity associated with Leica Geosystems.

“Looking at how far we have advanced, the CityMapper hybrid sensor now reaches a new tier in its third generation,” said Werner Kirchhofer, Vice President Product Development, Airborne Sensors Business Unit, Scanning & Mapping Division, Hexagon. “Our effort centered on higher performance and giving customers full freedom to set up the optical and laser payloads per mission, so one platform can span all scenarios while maintaining the Leica Geosystems standard of quality.”

Leica Citymapper-3 Airborne Lidar For Regional And Urban Mapping

A browser‑based utility makes it possible to generate, then deliver, an XML package, enabling rapid centralized configuration of GNSS options in Eos Tools Pro.

Across organizations, a leader in position technology has introduced the Eos MDM Configurator from Eos Positioning Systems. Built for the browser, the application lets teams rapidly build and roll out preset Eos Tools Pro configurations to large fleets of mobile devices through their MDM platforms, including options such as datum.

App Configuration With eos mdm configurator™

To replace hand‑scripting, administrators follow a guided flow for setup and only then see the XML rendering as they go. The sequence spans five areas related to satellite positioning, providing clarity before export for GNSS use. When the process completes, the export creates an output in XML that is ready for rollout to any external MDM. Through centralized controls, the organization can streamline setup and improve GNSS data consistency in Eos Tools Pro, while saving staff hours otherwise spent on per‑app tweaks.

Mdm Deployment For Eos Tools Pro: A Faster Way to Configure

Across Japan, a new rebrandable platform debuts for industry teams, aimed at local workflows, through a partnership between Pythagoras BV and Terra Drone.

Pythagoras CAD and Point‑Cloud Integration

At the core of the release, SLAM know‑how and airborne systems from Terra Drone are paired with Pythagoras tools for CAD work and dense point‑cloud handling, streamlining mapping tasks for surveyor teams and site crews. Delivered as a tailored edition of Pythagoras CAD, the package adds new tools and a workspace tuned to Terra Drone clientele.

Geospatial Impact on Survey and 3D Mapping

Beyond a simple rebrand, the move advances the geospatial field by marrying Terra Drone platforms with Pythagoras point‑cloud processing. The aim is clear: field and construction work gains higher throughput through quicker data handling and tight links with SLAM‑driven UAV operations; Pythagoras strengthens its role in configurable software, while Terra Drone expands its service depth.

Terra Drone Software For Uav Workflows in Japan

Following a rapid regulatory win, the venture moved to place the opening pair of telecom spacecraft into its own LEO system, securing high-priority Ka-band access only a week earlier after completing spectrum filings to expand secure connectivity worldwide.

From New Zealand’s Mahia Peninsula, Rocket Lab’s Electron carried two units on a ride branded ‘The Cosmos Will See You Now,’ marking the network’s switch-on phase for a scalable constellation intended to serve customers across Europe and beyond. Designed for future versatility, the platform can host earth observation sensors as well as telecom payloads.

At precisely 10:52 GMT on 22 January—11:52 p.m. NZDT the same day—the satellites separated into a circular orbit roughly 1,050 kilometers up, transitioning the program from factory and design work to in-orbit checkout.

Beyond pure engineering spectacle, this flight validates that the architecture, build pipeline, and operations approach are ready for routine service, paving a phased rollout of the wider system in the coming months to deliver resilient connectivity across Europe and the world.

Initial service targets range from Portugal to Liechtenstein.

Speaking after the deployment, Founder and CEO Rafel Jordà Siquier emphasized the strategic step.

This marks a landmark for our team and moves us closer to delivering trusted, sovereign broadband for Europe and partners worldwide. Progressing from spectrum authorization to operational spacecraft in mere days shows the maturity of our system and how quickly we can translate strategy into action.

These first craft establish the base for a robust space backbone designed to support governments, institutions, and commercial users when continuity matters most.

Open Cosmos Launches First Two Satellites to Low Earth Orbit

With confirmation from tracking teams, two company-built spacecraft are now in a round path about 1,050 kilometers high, following a clean deployment.

Following a rapid cadence in regulatory progress, the firm unveiled the first nodes of its proprietary LEO telecom constellation, arriving barely a week after it locked in high-priority Ka-band access. To advance global connectivity and insight for users, the team is building this network to carry communications at scale.

Satellite Mission and Constellation Details

As the evening unfolded on the East Coast of the North Island, the Mahia Peninsula in New Zealand hosted Rocket Lab’s Electron, which lifted two satellites under the banner “The Cosmos Will See You Now.” Together, the two form the opening move toward a future‑ready constellation built to provide scalable, robust, and synchronized services across Europe and farther afield.

Keeping to the planned schedule, the team recorded ignition at 10:52 on 22 January (GMT) — 23:52 NZDT — and, as the launch progressed nominally, the program moved from factory throughput and constellation design to on‑orbit checkout at an altitude near 1,050 kilometers on a circular path.

Beyond the headline moment, the event demonstrates that the architecture, build flow, and operations concept are ready to fly — a clear signal of mission success — and it sets up a phased expansion of the wider network over the months ahead, marking a milestone for deployment readiness.

Open Cosmos Launch Marks New Constellation Milestone

Across the country, a 3-centimeter program will be powered by three UltraCam Osprey 4.2 units acquired by Kavel 10.

Vexcel Imaging and UltraCam Osprey 4.2 Systems

In a recent update, the manufacturer confirmed that the Dutch provider has expanded its fleet with two Standard bodies and one Tele variant of the large-format camera from Vexcel Imaging to underpin a countrywide survey. Additionally, a full photogrammetric chain is part of the deal—UltraMap handles processing through orthoimage generation, enabling high automation.

Oblique Perspectives and 3D City Modeling

Through the combined camera-and-software stack, the company intends to photograph the entire country at a GSD of 3 centimeters. According to the company’s co-CEO, the initiative raises the bar for local airborne cartography.

From one harmonized dataset we will deliver a true orthoimage, high‑grade oblique images, and a richly detailed 3D model—providing the clearest possible view of the Netherlands and creating a scalable base for our long‑term plans,” said Richard Brouwer.

UltraMap Workflow and Aerial Surveys Efficiency

Operationally, the new camera system trims hours aloft by roughly one‑third, and all collected imagery is processed in a single, unified UltraMap pipeline. This integrated approach speeds delivery of geospatial data while protecting image quality.

Kavel 10 Adds Three Ultracam Osprey 4.2 Systems For Countrywide 3 cm Mapping

Through a shift to in-space compute, PCI Geomatics (DBA CATALYST) unveils a capability that outputs precision-corrected imagery aboard the vehicle, accelerating insight for operators. By moving heavy workloads upstream, missions allocate power and CPU more wisely while preserving imaging quality. The approach brings photogrammetric rigor to the edge without sacrificing fidelity.

Instead of beaming unrectified frames for later correction to faraway analysts, operators can receive geo-ready maps that reach Earth already corrected. The earlier ground-heavy model demanded expansive facilities, high‑rate links, and specialty staff, stretching timelines. Distortions from terrain and the sensor model are corrected before delivery. Within this flow, the step list includes terrain rectification, and only after several passes does orthorectification happen upstream of transmission.

The legacy path also leaned on large antenna farms and compute clusters, which drove up costs and turnaround time. As fleets continue to grow, scarce spectrum must serve many modern platforms in space, which lengthens queues. When capacity is rationed among many vehicles, the effect is amplified for large satellite platforms.

Inside the CATALYST software chain, two core operators—responsible for terrain correction and precise positioning—run in a lean in‑space pipeline tuned for modern rad‑hard processors. Because compute happens prior to sending data home, delay is cut drastically and insights flow while events are unfolding. With this approach, operators act immediately during time‑critical scenarios, including on‑orbit detections and rapid tasking. Applications range from disaster response to monitoring infrastructure, with algorithm support for onboard analytics. Because compute now runs aloft, a cascade of corrections executes first—map alignment, sensor model updates, and only then orthorectification and geolocation—so users act faster.

During emergencies, a cloud and shadow filter can run on multispectral collections before the pass ends, pushing alerts in minutes and cueing aerial assets.

Orthorectification at The Source For Satellite Missions

Editor's Note: In a field that evolves as rapidly as geospatial information science and technologies, the idea of “getting a GIS job” may not be as straight-forward as it sounds. What are employers looking for, and how do you know that your training and education will get you there? Join Directions Magazine as we begin a short series of articles examining these topics.

Was there really ever a time when all you needed to know to get a GIS job was how to do a few software tasks and design a map or two, or is that as overly idealistic and unrealistic as the image of every 1950s household having an apple pie cooling on an open window sill?

If you did manage to get a job on that "lick, spit and promise," are you still in that position? What daily tasks are expected of you now that didn’t even exist 15 years ago?

New GIS degree and certificate programs continue to launch each year. Keeping curricula current, and instructors both confident and competent, is a perpetual challenge. There are also more numerous and diverse approaches to professional development than ever before. Is there validity in the Monday-morning-quarterback statement that what employers want, what degree programs provide, and what students actually learn are always out of synch?

In an effort to understand distinctions and requirements of GIS jobs, Jung Eun (Jessie) Hong, an assistant professor in the Department of Geosciences at the University of West Georgia, recently completed a content analysis of GIS job advertisements. She compiled almost 1000 GIS job postings, spanning 2007 – 2014, from GIScareers.com, GISjobs.com and the GIS Jobs Clearinghouse. The titles of the jobs were used to sort the positions into five different job categories:

Though programmers, developers and engineers differ in training and expectations, their primary job tasks are similar enough to be grouped at this scale. The “Other” category included GIS coordinators, instructors and the like.

Then, individual skills specified in each ad — such as data mining, web mapping, programming or project management, for example — were all coded into four technical areas:

Hong then used this coded information in NVivo, a product designed for analyzing text-based content.

What were the similarities and differences in skills among the five job categories? The most requested skill set across all jobs involved analysis and modeling, with over 56% of all job ads specifically referencing such needs. This includes such tasks as aerial image interpretation, data analysis, database development, data mining, network analysis and/or the use of spatial statistics. The second most requested skill set was a general one: communication, interpersonal skills, and self-motivation and independence.

Within job categories, the similarities were more interesting than the differences. Not surprisingly, the number one most requested skill for a programmer/developer/engineer was web or mobile application development; 57.4% of all job postings within that group identified that skill. But what was the most frequently expected skills for the other four categories of GIS jobs positions? Communication skills, above all other technical or more job-specific needs, with up to 45.9% of the postings specifying that skill set.

Another revealing result from within each job category was that database development was ranked as the second most frequently desired skill across all five job type categories. Within the domain of GIS&T, "database development" can have a specific meaning for a specific use case, but those two words are also common and important on their own within the fields of computer science and information technology. Their high frequency designation as a skill across all of these GIS job postings may reflect diverse and different interpretations of what “database development” actually means in a given scenario...which makes us wonder about the myriad ways that “communication” itself plays out in the world of GIS&T.

In practice, communication skills can include everything from accurately representing one’s self on a resume to finding out during an interview that everyone is interpreting “database development” in distinctive ways. In the Venn diagram world of database development, surely there is overlap with data manipulation, programming, and/or database design. Are you ready for that? Or, does the new boss really just need someone to populate an Excel spreadsheet and join it to a shapefile? Either way, whether you are a wise boss or an eager job seeker, are you prepared to listen carefully enough to the other person so you could tactfully, professionally and courteously adjust and address the miscommunications? Voila, now you get points for interpersonal skills as well!

If the ubiquity of “database development” may reflect instances of commission, Hong’s research also illustrates an example of omission with the absence of the term “geocomputation” in any of the job ads. She had been prepared to code the term as an example of a required or desired skill within the analysis and modeling set, based on the fact that geocomputation had been identified as an entire knowledge area in the 2006 GIS&T Body of Knowledge, one of the sources for her coding information. Geocomputation had been described in the BoK as the "development and application of computationally intensive approaches to the study of complex spatial-temporal problems.” The complete absence of the term from job ads does not mean that those complex spatial-temporal problems are no longer an issue! Instead, I would suggest that the high performance computing, cellular-automata and agent-based models, and simulation modeling that once differentiated a geocomputational approach have now become expected and necessary, and thus have been integrated into analytical approaches in general. Perhaps “Big Data analytics” is more likely to be the nom du jour in a job ad of today, but these are of the same ilk.

Overall, Hong’s findings confirm what many of us have experienced over time: that technical skills will always be critical on a particular day for a particular task, but those must be complemented with a life-long ability to unravel problems. When the Department of Labor’s Geospatial Technology Competency Model was revised in 2014, the bulk of the changes were specific technical competencies in Tier 4 and above. Personal effectiveness, academic and workplace competencies, largely the "general" skills in Hong’s study, have remained as solid and important as ever.

In psychological and human resources parlance, skills are things which we can acquire and learn, while abilities are things that we have naturally. Importantly, both can be refined through training and education. To increase your employability, hone your skills as well as your abilities, and develop your competence and confidence to communicate about them both. Consider these 20 Challenging GIS Interview Questions, which are as relevant today across all types of GIS positions as when they were first published; effective communication is central to them all. 

GIS Jobs of Today: Which GIS job skills are in highest demand?

Geospatial technology has become a fundamental component of modern information systems. As societies generate growing volumes of data, the ability to organize, analyze, and interpret information through geographic context has transformed how organizations understand the world around them.

Mapping is no longer limited to visual representation. Today, geospatial tools support analysis, prediction, and strategic planning across a wide range of industries.

From Geographic Data to Spatial Understanding

At its core, geospatial technology connects data to place. This connection allows users to explore relationships between location, time, and human activity. By layering multiple datasets within a geographic framework, analysts can identify trends and interactions that are difficult to detect in non-spatial data.

Directions Media

Press Releases

Read more