Asking for an Image Sequence

An image sequence is a sequence of images, similar to an animation, with the exception that each image in the same sequence might have different characteristics such as geographic bounding box, image height, and image width. This type of animation is useful when tracking a particular phenomena, such as a hurricane, across a wide area without serving huge images or losing image resolution.

An image sequence is advertised in the Capabilities XML by a named layer that also contains a new Dimension or Extent tag named "sequence". The child layers of this named layer contain all of the pertinent metadata for each image in the sequence, and are the preferred source of metadata. However, in order to preserve the concept of each named layer representing an animation that varies in one or more dimensions, the differentiating metadata can be obtained directly from the "sequence" Dimension or Extent tag.

As an example, here is an excerpt from the Capabilities XML (WMS Version 1.3.0) describing Hurricane Isabel from 2003 (some tags have been removed or edited for clarity):

<Layer opaque="0" noSubsets="1" fixedWidth="1024" fixedHeight="1024">
  <Name>2919_21239</Name>
  <Title>Hurricane Isabel (Sequence)</Title>
  <Abstract>This sequence of images was used to create an animation 
    of the progression of Hurricane Isabel as seen by MODIS.</Abstract>
  <BoundingBox CRS="CRS:84" minx="-90.356714749713674" 
    miny="7.609286726629172" maxx="-33.926345696206447" 
    maxy="50.290073918712146"/>
  <Dimension name="sequence" units="" 
    default="TIME(2003-09-18T15:55Z)XMIN(-89.936816774797592)YMIN
    (20.999591279609998)XMAX(-58.935329635126976)YMAX(50.290073918712146)
    WIDTH(2048)HEIGHT(2048)">
      TIME(2003-09-08T13:45Z)XMIN(-50.709066462147263)
      YMIN(7.609286726629172)XMAX(-33.926345696206447)
      YMAX(26.238583752241105)WIDTH(2048)HEIGHT(2048),
      TIME(2003-09-10T16:40Z)XMIN(-61.935721735776916)
      YMIN(12.634935137084874)XMAX(-40.779393874440224)
      YMAX(29.015256506204437)WIDTH(2048)HEIGHT(2048),
      TIME(2003-09-18T15:55Z)XMIN(-89.936816774797592)
      YMIN(20.999591279609998)XMAX(-58.935329635126976)
      YMAX(50.290073918712146)WIDTH(2048)HEIGHT(2048)
  </Dimension>
  <Style>
    <Name>overlay</Name>
    <Title>Transparent Overlay Only</Title>
    <Abstract>This is a composite image of all of the individual frames 
      superimposed on each other.</Abstract>
  </Style>

The first few lines look like other images and animations provided by the SVS Image Server. However, there is a new Dimension tag whose name is "sequence". This indicates that this layer represents an image sequence where each image does not have the same geographic bounding box. This Dimension does not have any units, has a default value, and has three valid values. Each value is comprised of several sub-strings of the form "key(value)". Each value is a long string without spaces or newlines, while they have been broken up on different lines in this example for clarity. Different values are separated by commas. From the example, the default image has a timestamp of "2003-09-18T15:55Z", 2048 as its width, 2048 as its height, and the geographic bounding box of (-89.936816774797592, 20.999591279609998, -58.935329635126976, 50.290073918712146).

When making a request for a particular image in this sequence, a client should substitute the values within the "sequence" into the corresponding values in a normal WMS request, as well as passing the value of the "sequence" dimension unchanged. For example, to request the first image in this sequence, the proper request is:

http://svs.gsfc.nasa.gov/cgi-bin/wms?REQUEST=GetMap&VERSION=1.3.0&
LAYERS=2919_21239&CRS=CRS:84&FORMAT=image/png&BBOX=-89.936816774797592,
20.999591279609998,-58.935329635126976,50.290073918712146&HEIGHT=2048&
WIDTH=2048&SEQUENCE=TIME(2003-09-08T13:45Z)XMIN(-50.709066462147263)
YMIN(7.609286726629172)XMAX(-33.926345696206447)YMAX(26.238583752241105)
WIDTH(2048)HEIGHT(2048)

The use and interpretation of the "sequence" Dimension is entirely optional. Each individual image can be requested from the child layers of this named layer and then played in sequence to show the complete animation.

When this named layer is requested without the "sequence" dimension, an overview image that contains all of the individual images composited together is returned. The Bounding Box value, the fixedHeight and fixedWidth attributes in the layer all refer to the overview image. Here is a sample request to obtain the overview image:

http://svs.gsfc.nasa.gov/cgi-bin/wms?REQUEST=GetMap&VERSION=1.3.0&
LAYERS=2919_21239&CRS=CRS:84&FORMAT=image/png&BBOX=-90.356714749713674,
7.609286726629172,-33.926345696206447,50.290073918712146&HEIGHT=1024&
WIDTH=1024

Go backward to "Asking for an Animation"

Go forward to "Design Considerations - Fixed Size Images"