Overview | Plot Types | Examples of Using SAGE for Scientific Analysis |
---|
The Services for Analysis of the Greenland Environment (SAGE) is a tool designed to help scientists access, integrate, and analyze data related to the history and status of Greenland's ice sheet in real time using timeseries, scatterplot, histogram, and box and whisker plots. Currently SAGE supports seven data sets. See Table 1 for a listing of those data sets.
DIF ID |
Data Set Title |
---|---|
NSIDC-0032 |
DMSP SSM/I Pathfinder Daily EASE-Grid Brightness Temperatures |
NSIDC-0066 |
AVHRR Polar Pathfinder Twice-Daily 5 km EASE-Grid Composites |
NSIDC-0218 |
Greenland Ice Sheet Melt Characteristics Derived from Passive Microwave Data |
NSIDC-0301 |
AMSR-E/Aqua Daily EASE-Grid Brightness Temperatures |
NSIDC-0305 |
GLAS/ICESat 1 km Laser Altimetry Digital Elevation Model of Greenland |
NSIDC-0342 |
Near-Real-Time DMSP SSM/I-SSMIS Pathfinder Daily EASE-Grid Brightness Temperatures |
MOD11A2 |
MODIS/Terra Land Surface Temperature & Emissivity 8-Day L3 Global 1 km SIN Grid V005 |
The SAGE interface is separated into five main screens: What, Where, When, Search Criteria, and Current Workspace.
On the What screen, you choose the data sets and the variables you are interested in researching. This screen functions as a collapsable tree to make viewing and choosing of the data sets and their variables easy.
On the Where screen, you specify the spatial coverage you are interested in by using a location map. There are three ways you can do this:
On the When screen, you choose the temporal coverage for your data with the option to specify annually recurring interval data. The Intra-annual Range field is where you can refine your data search even further by entering either the month and day of month (MM-DD), or the Day of Year for a given search.
The Search Criteria screen allows the user to review and manage all current selections.
The Workspace screen contains results of all operations: search, analysis, and plot, and it is persistent between sessions (meaning through cookies on the same computer using the same browser the data will be accessible for 14 days), allowing you to come back later to get the results for plots that take a long time to generate.
SAGE can create timeseries, scatterplot, histogram, and box and whisker plots. Below are examples of each type of plot SAGE can create.
These are thumbnail images for each time step in a series. The number of thumbnails per page is configurable.
A circulation index, such as El Niño/Southern Oscillation, can be added to a time series plot.
The threshold vales used in creating these averages can be specified by the user.
A time series plot can display up to five different variables over the same time interval.
A histogram is displayed for each time step in a series.
SAGE is a powerful tool in that it allows the user to find the data and then analyize it wihout leaving the interface. In order to help our users understand the enormouse scope of this tool, we have offered a scenario of how a user may use SAGE for scientific research.
A crysopheric scientist is interested in how the mass balance of the Greenland Ice Sheet is changing. She wants to explore the spatial and temporal patterns of surface melt on Greenland and investigate how these are related to other surface properties such as albedo and surface temperature, as well as atmospheric circulation patterns. She is aware of some relevant data sets but is uncertain where to find them, and she wants to do her exploratory work remotely without having to locate, decode, co-register, and analyze these heterogeneous data sets.
She begins by selecting the Greenland Melt data set and the drainage basin for Jakobshaven glacier. She doesn't give a time range, and so gets 4233 hits.
She requests a time series over the entire time period for the Jakobshavn glacier, which produces a plot showing an upward trend of 0.399 melt days per year. Refer to Figure 1.
Figure 1. Time Series Over the Entire Time-period for the Jakobshavn Glacier
She notes that 1998 has the greatest melt days of the record while 2000 has a minimum so she decides to look more closely at these years.
She first plots a time series of the AVHRR albedo and the AVHRR surface temperature for the same time period to see how surface conditions vary. Refer to Figure 2.
Figure 2. AVHRR Albedo and Surface Temperature Time Series Plot
She then selects the SSM/I 19 Ghz horizontal polarity ascending brightness temperature with the AVHRR albedo and plots them together as a time series. Refer to Figure 3. She does the same with the brightness temperature and the surface temperature to see if there is a relationship between albedo and temperature. Refer to Figure 4.
Figure 3. SSM/I Brightness Temperature and AVHRR Albedo Comparison Time Series Plot |
Figure 4. SSM/I Brightness Temperature and AVHRR Surface Temperature Time Series Plot |
---|
Curious to see how the melt pattern in this basin varies with the NAO, she filters the SSM/I brightness temperature based on a NAO of ±0.5. Refer to Figure 5.
Figure 5. SSM/I Brightness Temperature Based on a NAO of ±0.5
She then picks a drainage basin on the eastern side of Greenland to compare and contrast melt and surface temperature patterns for 1997-2000. Refer to Figure 6. She then creates a comparison time series plot of the AVHRR albedo and the AVHRR surface temperature to SSM/I 19 Ghz horizontal polarity ascending brightness temperature. Refer to Figure 7.
Figure 6. Greenland Ice Sheet Melt Characteristics Derived from Passive Microwave Data | Figure 7. SSM/I Brightness Temperature and AVHRR Surface Temperature and Albedo Comparison Plot |
---|
Interested in the patterns she sees, she decides to do a small image time series of the melt extent, the surface albedo, and the surface temperature. Refer to Figures 8, 9, and 10.
Figure 8. Greenland Ice Sheet Melt Extent from SSM/I Data
Figure 9. AVHRR Surface Albedo Comparison Plot
Figure 10. AVHRR Surface Temperature Comparison Plot
Finally, she selects a point near the summit of the ice sheet and points near the outlets on the western and eastern sides. She then does the same temporal query on the MOD11A2 data set for each, and then she plots a time series of the land surface temperature at each point. Refer to Figures 11, 12, 13 and 14.
Figure 11. Summit MODIS/Terra Land Surface Temperature/Emissivity 8-Day L3 Global 1 km SIN Gird V005 (Mean) Plot | Figure 12. Western Coast MODIS/Terra Land Surface Temperature/Emissivity 8-Day L3 Global 1 km SIN Grid V005 (Mean) Plot |
---|---|
Figure 13. Eastern Coast MODIS/Terra Land Surface Temperature/Emissivity 8-Day L3 Global 1 km SIN Gird V005 (Mean) Plot | Figure 14. Comparison Time Series (Mean) Plot |
She downloads the data sets of interest and does further investigation with tools on her desktop. Later she decides to investigate the years 2001 and 2002, also contrasting in terms of number of days of melt, she downloads additional data such as DEMs and accesses the updated data via NSIDC's SubsetAgent/GetData API.
None Selected
None Selected
None Selected