Some of you might know PATN already, but for those amongst you who don’t, PATN does pattern analysis. It is comprehensive and extremely versatile and excels at extracting and displaying patterns in any type of complex (multivariate) data, yet I found it simple to use.  PATN is different from statistical packages such as SPSS or SAS as it specialises in the exploratory aspects of statistics (such as pattern finding and hypothesis generation). The results produced by PATN can then be used in any statistical software to do hypothesis testing.

So why would you do pattern analysis?

• If you can identify patterns in your data you can actually start to understand the variation and the processes that cause that variation in your dataset, and
• PATN helps you to locate errors in your data, to summarise complex information, to create an independent perspective of what your data tells you and also to reduce noise in your data.

In Day 1 of the class we covered lots of theory to create a basic understanding of how PATN works and to then being able to use it appropriately. This included the entire process of how to analyse data in PATN. We also learned about intrinsic and extrinsic variables, variable scales, transformation and standardisation, association measures, hierarchical and non-hierarchical classification techniques and much more until our heads were steaming. Whilst I know these terms from my GIS and Remote Sensing background PATN still requires a quite different approach of thinking which is why this class was a big learning curve for me.

Day 2 was more hands on. Many of us had brought in our own data to analyse and so that’s what we did. For some of us this was pretty straight forward and first results were quickly produced. For others however, the analysis was a bit more complex and problems in the data sets became obvious. This was a really good outcome as it helped people to gain a better understanding of their data which really is what PATN is all about: understanding, interpreting and evaluating what your data is telling you. Also, a really nice group dynamic developed during the day and those attendees among the group with more experience helped out the beginners.

Although this was a strenuous two days we made sure to stay on top of it all by regularly regaining energy from the local coffee suppliers and having lunch at Siena’s each day.  Debbie and Tom did a lot of running around to make sure this all ran smoothly. And of course Lee did a fantastic job of teaching us so much information in just two days, and really giving us a great start in using PATN.

Gaia Resources hopes to run more of these courses in the future, so keep an eye out for more details or drop us a line if you are interested and we’ll organize another course.

Feel free to leave us a comment below or email me directly,

Grit

The post PATN Training Course appeared first on Gaia Resources.

The first thing we did was a requirements analysis to find out what exactly our client needed. After a review, we realised that we did not need to acquire any new data, as all this had been collected by our client already for other purposes. Given the data was in hand, we started work very quickly on the methods. As there were a large number of waste dumps across a large area for which the volume of materials had to be calculated, we decided that using the available data in an automated method would be the best option for the client.

A subset of the mine site showing the aerial imagery

The data we had consisted of the waste dump polygons for various mine sites (see examples in the images below), NearMap photo maps as well as a tiled NearMap digital elevation model (DEM). Our preparation of the data for further processing included:

• The creation of a location index for the DEM tiles to identify those tiles that covered the actual mine sites,
• Creating mosaics of DEM tiles wherever waste dumps were located across several tiles (as below),
• Re-sampling the DEM tiles to a higher resolution using a nearest neighbour algorithm to closely fit the raster cells of the DEM to the waste dump polygons, and
• Deriving the base level elevation of each waste dump.

Converting DEM tiles to mosaic

The base level elevation was defined for each group of connected waste dumps (as below) by extracting the elevation at a number of points located close to the outside border of the waste dump polygons.

Calculating the base level around the waste dump polygons

For the actual volumetric calculations we assembled a number of Spatial Analyst tools (map algebra as well as zonal statistics) in a python script to automatically calculate volumes for each waste dump. These volumes were then used to derive waste dump type dependent material balances.

To account for the angled side walls and the subsequently different sizes of the bottom and the top areas of a waste dump, volumetric calculations needed to be carried out on a pixel by pixel basis as opposed to using the entire waste dump area. The resulting output was a table listing each waste dump and associated volumes of different types of materials such as topsoil and capping material.

This project was a great opportunity for us to be creative and develop a new method that delivers an easy solution to a problem we’ve not dealt with before – and solve our client’s problem for them.

If you have any questions or comments, feel free to email me, leave a message in the comments below, or start a conversation with Gaia Resources on Twitter.

Grit

The post Volumetric calculations for waste dumps appeared first on Gaia Resources.

Gaia Resources staff have volunteered to set up the company’s Anabat SD2 detector at their homes to record bat calls for two to three consecutive nights in a row on an alternating schedule. Location specific information is recorded during this time which includes traffic volume, special events that may have occurred (thunder storms, sporting events, concerts etc.) and the exact location outside the house where the Anabat was set up. Weather information is sourced from the Bureau of Meteorology (BoM), aside from this, information such as moon phases and sunrise/sunset times are also recorded.

Identification of bat species is currently underway (with a lot of support from Kyle Armstrong from Specialised Zoological), the process itself is time consuming but despite this we hope to have it complete by the end of the year. During this process we have discovered that five different bat species out of the eight occurring species, frequent the Perth metropolitan area; this preliminary result is based on 250 individual bat calls already identified. Three of the bat taxa that we have recorded are Chalinolobus gouldii or Gould’s Wattled Bat, Mormopterus sp. or Freetailed Bats (we’re still working on identification to species level) and Tadarida australis or the White-striped bat.

Chalinolobus gouldii, one of the bats we record regularly

All data collected during our bat survey is being entered into the Gaia Resources Biological Data Recording System (BDRS) implementation; this data includes .DAT files of bat calls, species identification and ancillary information.  The BDRS has been the subject of many blogs, Piers’ own birding records are kept in there (over 11,000 records now), as well as previous data we collected on trees.  This bat research has also contributed to much testing and use of the BDRS by the rest of the team – and that helps our software engineering staff to spot issues and iron out bugs in the system.

One of the things we discovered during our urban bat survey is the influence of ultrasound noise surrounding us in the urban environment. Sources of ultrasound noise can be air conditioning systems or computers. The Anabat will pick up these ultrasounds, and in turn produces large amounts of data. Bats may also avoid areas of ultrasound noise as it can interfere with their echolocation, so optimisation of Anabat setup (sensitivity and angle) and placement of the Anabat was something we investigated and played with to attain the best results.

In the near future, Debbie and I will complete the identification of species from all bat calls recorded (with Kyle’s help). Once this is done, the data will be analysed focusing on a number of questions such as bat abundance and species presence at different locations and their association with environmental factors in the closer and further surroundings of the recording locations.

Thus far, the Gaia Resources urban bat research project has been a rewarding and informational process and we look forward to continuing on with it and gaining more knowledge in the world of bats. All of the idea and work for this project started well before Piers and I went batty in Melbourne, but that conference certainly kept us interested and intrigued in continuing this research.  For more information on bats in Australia, New Zealand and Papua New Guinea visit the Australasian Bat Society.

Feel free to email me, leave a message in the comments below, or start a conversation with Gaia Resources on Twitter.

Grit

The post Perth metropolitan bat survey appeared first on Gaia Resources.

Remote Sensing (RS) optical and radar data is widely used to create spatially coherent data sets of earth surface and atmospheric properties at local to global scales covering large areas. On the contrary, field measurements are usually spatially restricted. Some examples of RS data sets are climatic variables (e.g. surface temperature, humidity and precipitation), land surface properties (e.g. vegetation cover, type and condition, geology, soil moisture, inundation, salinity, terrain and human interactions).  This is an area that Gaia Resources has been working with for some time now, and we thought it might be useful to provide a bit of an overview as to the capabilities as we see them.

RS data present an important resource for various applications e.g. the mapping of anthropogenic and natural changes, species distribution and habitat modeling, climate change and human impact assessment, mineral exploration and disaster management. There are many applications in agriculture and forestry that rely on RS data.

The fact that RS data is captured repeatedly (at least in the case of satellite data) covering vast as well as remote areas that are difficult to access makes its usage particularly attractive for many areas in Australia which is physically large but very sparsely populated and do therefore suffer from a lack of (temporally regular) ground measurements.

A wide range of freely available satellite data is distributed by NASA’s Earth Observing System Data and Information System (EOSDIS) data centers. These distribution centers provide different online search and order tools such as the USGS Global Visualisation Viewer GloVis, the USGS Earth Explorer (EE),  the Warehouse Inventory Search Tool (WIST), or others that can be found on this website. A limited number of satellite data sets for Australia is being made available through Geoscience Australia.

The different portals offer various types of data sets from different sensors. These data do not only have different properties regarding their spatial, temporal and spectral resolution which determine the field of application of the data but there are furthermore different processing levels of each data sets. These processing levels range from raw radiometrically, geometrically and atmospherically uncorrected data to higher level products that have been corrected and contain derived information such as vegetation indicies. These two websites from the Goddard Earth Sciences and Information Services Center and the Land Processes Distributed Data Active Archive Center give an overview on some of the data sets available.

Products from the Moderate Resolution Imaging Spectroradiometer (MODIS) and Landsat sensors are popular freely available data sets. MODIS is a key instrument aboard the Terra (EOS AM) and Aqua (EOS PM) satellites launched in 2000 which is viewing the earth surface every 1 to 2 days acquiring data in 36 spectral bands, or groups of wavelengths at a spatial resolution of 250 metres to one kilometer. These data are particularly useful to frequently monitor large areas over longer time periods. Some examples of information derived from MODIS data of different Australian regions are NDVI images visualizing a multi-year draught in eastern Australia, images of the 2011 fires in the Northern Territory or the 2010/2011 flooding in Victoria.

Landsat records data at a higher spatial resolution of up to 15 metres making it suitable for local scales but has a much lower revisit rate of about 16 days. The current sensor, Landsat-7 of the Landsat family launched in 1999 provides 7 spectral bands which makes Landsat-7 imagery ideal for vegetation, water and land monitoring. However, one Landsat scene covers a much smaller area of approximately 170 km north-south by 183 km east-west which implies that more scenes and therefore greater data processing efforts are needed for larger areas.

More and more data products providing higher spatial resolution are becoming commercially available. Examples are the WorldView Satellites and GeoEye which currently provide the highest spatial resolution available. Other sensors are for example Quickbird, Ikonos, SPOT, ALOS and ASTER.

Gaia Resources can provide a range of services to assist you with your remote sensing needs. You can follow up with me on this article by leaving a comment below or sending me an email.

Grit

The post What could Remote Sensing data do for you? appeared first on Gaia Resources.

We looked at a number of different commercial and open source Laboratory Information Management Systems with the aim of making it easy for one of our clients to capture data from their specific workflow. While there are a number of solutions available, some are too specific to capture alternative use cases, others were not maintained and only two open source options came to the fore. Jump to the summary to find the conclusions and find out how to give us feedback.

Background

A number of our clients are starting to want to capture and report on the results of different experiments (treatments) on entities in their collections, and to that end we have decided to look at the available existing solutions. The particular use case that they have in mind is Seed Banking and capturing the results of various tests such as germination tests. In the context of this we have been reviewing different data management solutions. One particular solution that came up in our review is the concept of a Laboratory Information Management System (LIMS).

What is a LIMS?

It is a software based tool that helps support laboratories and the workflows performed within them. Their key features are workflow and data tracking support, flexible architecture, and smart data exchange interfaces.

Who uses a LIMS?

People who are tired of capturing laboratory data in myriad seperate notebooks, spreadsheets, text file outputs and who believe there must be a better way of using modern technology for their benefit rather than to their detriment.

Requirements

We are looking for a tool that will capture information about see Table 1 below for a list of features that a typical LIMS has.

Table 1 Typical LIMS functionality (taken from (Paszko & Pugsley 2000)⁠)

On top of the features above my review included the following considerations:

Table 2 Additional considerations in selecting an appropriate LIMS.

Available LIMS

The list of available LIMS was taken from a number of sources including the LIMS Book and Buyers Guide (Jones & Vaughan 2006)⁠, Benn and Liscouski 2009⁠, an Open Source LIMS Wiki Page (Anon 2009)⁠, a list on Sourceforge (Sourceforge 2011)⁠ (Institute for Laboratory Automation n.d.)⁠ and a review of results using Google Scholar.

Table 3 List of LIMS reviewed

Review

A data matrix of the different LIMS was prepared, and is embedded below or can be found on google docs.

A summary of these can be found in Table 4 below.

Table 4:Coverage of features by the different LIMS

Recommendations

From the above review we would recommend that if both cost and extensibility are the main issues at hand, LabKey would be easiest to implement and customise to the client’s need while GRIN-Global could possibly be investigated in terms of the client’s workflows. Bika is another strong open source LIMS that would warrant investigation, although it’s reliance on Zope’s ZOPEDB rather than a generic RDBMS could be it’s downfall.

References+

The above review is by no means exhastive. If you know of any solutions that coudl be added to this list then contact me via email or twitter or leave a message below. There’s also a group on Mendeley where we’ve kept track of any additional references relating to this field – feel free to request access to the Mendeley group or the google docs spreadsheet.

The post Systems and How Useful They Are appeared first on Gaia Resources.

Today is also the first day for Tom Brownlie, who joins us as a Senior Consultant.  Our People page will be updated soon with information about both Tom and Anthony.

The post New staff and Wikimedia recap appeared first on Gaia Resources.

The post Conferences? appeared first on Gaia Resources.

http://www.usgs.gov/newsroom/article.asp?ID=2076&from=rss

The bit that’s exciting:

“In a breakthrough applauded today by the international Group on Earth Observations, scientists and decision-makers will soon have unrestricted global access at no charge to the USGS Landsat archive, the world’s most extensive collection of continuously-acquired land imagery. By the end of this year, the full collection will, for the first time, be freely available online to users around the globe…”

It’s not quite the game-changer that Google Earth was, but it is some good financial news in the current climate.  We’ve been working with Landsat data for a while and this might help kick off a few projects that had limited budgets in the new year…

Speaking of interesting economic times, there was an interesting podcast on Directions Magazine about the effect of the “downturn” on GIS.  It’s here http://www.directionsmag.com/article.php?article_id=2942 – download it from there in your favourite format if you are interested.  It’s very American centric but it’s quite optimistic about the next year for GIS.

Email me directly here.

The post FREE Landsat appeared first on Gaia Resources.