ENVI Unsupervised Classification Workflow

Software used: ENVI 5.1

The ISODATA method for unsupervised classification starts by calculating class means evenly distributed in the data space, then iteratively clusters the remaining pixels using minimum distance techniques. Each iteration recalculates means and reclassifies pixels with respect to the new means. This process continues until the percentage of pixels that change classes during an iteration is less than the change threshold or the maximum number of iterations is reached.

1. Start ENVI.
2. From the Toolbox, select Classification > Classification Workflow. The File Selection panel appears.

     

3. Click Browse. The File Selection dialog appears.

Advertisements

ESRI Launches New Site to Find Open Data

Search and Discover Authoritative Information from Any ArcGIS Open Data Website

ESRI today announced the launch of a new site aimed to help citizens discover organizations sharing open data around the world and provide direct access to thousands of open government datasets. Citizens can search, download, filter, and visualize this data through their web browser or mobile device.

Since July 2014, more than 1,200 organizations from all levels of government, including the National Geospatial-Intelligence Agency (NGA), and the cities of Raleigh, North Carolina; Tampa, Florida; Charlotte, North Carolina; and Muroran, Japan, have used Esri’s ArcGIS Open Data to configure custom open data sites to serve local citizens and businesses. Now the public can search across all these sites to find authoritative data by location and topic.

“We are excited about the large number of organizations currently sharing open data and believe we have a great opportunity to boost global support for open data and open knowledge,” says Andrew Turner, CTO of Esri’s DC R&D Center. “As more of the 380,000 organizations we work with across the globe begin to contribute open data, we will be able to help foster innovation by connecting the millions of datasets created by government agencies and shared through ArcGIS Open Data.”

Any organization can make its data available through ArcGIS Open Data, and people can now discover this data by visiting opendata.arcgis.com.

Learn more about ArcGIS Open Data at esri.com/opendata.

Source: http://www.esri.com/esri-news/releases/15-1qtr/esri-launches-new-site-to-find-open-data

Surface Temperature from Landsat Data: a New Lab Using the Semi-Automatic Classification Plugin

A very detailed tutorial for deriving surface temperature from Landsat data using QGIS and Semi-Automatic Classification Plugin.

Source: From GIS to Remote Sensing: Surface Temperature from Landsat Data: a New Lab Using the Semi-Automatic Classification Plugin.

Download the tutorial pdf (link)

In a previous post I have illustrated how to estimate land surface temperature using Landsat images and the Semi-Automatic Classification Plugin.

I was very pleased when Katie Fankhauser, a graduate student at Portland State University, informed me that she was preparing a lab, inspired by that tutorial, about how to determine ground surface temperature using satellite imagery and my plugin.

The document that she prepared provides background information about remote sensing and Landsat imagery (such as conversion of Landsat images to TOA reflectance and brightness temperature), and describes all the required phases about:

– download of software and data;

– data processing and supervised classification of land cover;

– calculation of an emissivity raster and estimation surface temperature.

Temperature calculated for a Landsat image of Portland (data available from the U.S. Geological Survey)

The image used in the lab is a Landsat image of Portland (OR, USA). The processing of data is described step by step, from the ROI creation to the raster calculation, which is ideal also for students that have little experience with remote sensing.

The lab document (pdf file) is freely available at the following link:
http://web.cecs.pdx.edu/~derekt/Classes/ME%20411%20Winter%202015/Laboratory/ME411_RemoteSensingLab4.pdf

This lab was prepared by Katie Fankhauser, and Evan Thomas who is Assistant Professor of the course Mechanical Engineering Measurements at the Portland State University, Sustainable Water, Energy and Environmental Technologies Laboratory. Very kindly, they credited also me as an author of this lab and allowed me to share the document.

Katie is currently working for a health campaign that aims to reduce the amount of wood fuel use consumed by traditional stone fires, and she is involved in ground truthing satellite-derived land surface temperature to study the rates of deforestation in Rwanda.

Evan Thomas holds a Ph.D. in Aerospace Engineering Sciences, and in particular he is the Director of the Sweet (Sustainable Water, Energy and Environmental Technologies) Laboratory, a very interesting and worthy project (www.sweetlab.org):

“At Portland State, the SweetLab designs and tests sustainable life support technologies for spacecraft and developing countries. The SweetLab’s current primary focus is developing and implementing remotely accessible instrumented monitoring technologies designed to improve the collection of effectiveness evidence in global health programs, including high efficiency cookstoves, water pumps, household water filters, sanitation systems, pedestrian footbridges and other developing world appropriate technologies. The SweetLab has projects in India, Nepal, Indonesia, the Philippines, Rwanda, Kenya, Uganda, Haiti and other countries with partners including the Gates Foundation, USAID, Mercy Corps, the Lemelson Foundation, the Global Alliance for Clean Cookstoves, and DelAgua. The SweetLab also has on-going work with the NASA-Johnson Space Center on microgravity fluid management systems”.

I am really glad and honored that my work can be useful for courses about environmental sustainability like this. Also, I hope that the upcoming new version of the plugin will allow for improved environmental analyses.

I would like to thank very much Katie and Evan and I hope that there will be other opportunities of cooperation in the future.

If one is interested in sharing the work done using the Semi-Automatic Classification Plugin, please contact me at the Facebook group or the Google+ Community.

2014 Most Cited Chinese Researchers

Congratulations to Prof. Xia Li, who is reported to be the most cited Chinese researcher in Social Science in 2014. The statistical data is based on the Scopus database from Elsevier.  Prof. Xia Li is an expert in the field of GIScience, remote sensing, cellular automata, agent-based modelling, and spatial optimization. He has published more than 200 peer-reviewed papers with a total of 6,683 citations in Google Scholar.  He is a Chair Professor at School of Geography and Planning of Sun Yat-sen University, the top ten university in China. I was so lucky to be have him as my supervisor when I was pursuing my master degree at Sun Yat-sen University.

Congratulation again to Prof. Xia Li for this big achievement!

Source: 2014年中国高被引学者榜单发布

NASA’s Soil Moisture Active Passive (SMAP) Mission Launches

The $916.5 million NASA’s Soil Moisture Active Passive (SMAP) mission has been successfully launched on January 31, 2015 from Vandenberg Air Force Base, at 9:20 am Eastern time atop a United Launch Alliance Delta II rocket. SMAP will acquire global soil moisture in every 2-3 days, providing data for weather forecasting, drought monitoring, flood prediction, crop productivity estimation, etc. Immediately after launch, SMAP enters a busy 90-day period called “commissioning.” It is expected that SMAP will start producing data in May 2015.

NASA SMAP webiste (link)

NASA SAMP Launch News (link)

NASA SMAP Launch Video (link)

Making the most detailed tweet map ever | Mapbox

An interesting article on geotagged tweets with open-source tools!

Making the most detailed tweet map ever | Mapbox.