(i)Parmiggiani F., (ii)Marra G.P., (ii)Quarta G., (ii)Conte D.
Institute of Atmospheric Sciences and Climate of the Italian National Research Council
(i)Via P. Gobetti, 101 - 40129 Bologna – Italy
(ii)S.P. Lecce-Monteroni km 1.2, 73100 Lecce – Italy
{f.parmiggiani, gp.marra, g.quarta, d.conte}@isac.cnr.it

Long abstract

1. Introduction

A first experience in developing a NOAA direct readout data management and presentation system, using 
both  commercial and Open Source components is described. The direct readout system is part of a ground 
station which receives, pre-processes  and archives data trasmitted by different remote sensing satellites. 
During its orbit and by means of its direct broadcast system, a satellite transmits real-time data to any 
ground station  within its fied of view. The direct readout of satellite data is performed by a system which 
consists of the antenna and of the telemetry system for tracking satellite passes and capturing raw data. After 
the acquisition, a pre-processing sequence is activated, which converts raw data to calibrated and geo-
referenced images.
Several commercial softwares are available to carry out the pre-processing phase and to archive the 
intermediate products to be utilized by geophysical scientists. Large scientific organizations, like NASA, 
NOAA, JPL, etc., have developed  web based infrastructures which provide advanced services for searching, 
viewing and accessing remote sensing products. Small scientific institutions are interested to develop low-
cost applications for data/product distribution with the aim to facilitate collaborative projects.
The objective of the system described in this paper is the  development of a web application which will 
improve access, viewing and retrieving of  remotely sensed data; the system is linked to the direct  readout 
station already operative in our Institute. With major details, the functions we are going to implement are:

	–	automatic downloading of remotely sensed data acquired by our direct readout station;
post-processing of the downloaded product for converting it in a common data format and processing level;
	–	extraction of a standard set of information from the downloaded product  (this will represent the 
metadata related to the product itself) and its archiving;
	–	allow the user to search the data on the catalogue using a graphical interface and using geographic 
search criteria;
	–	allow the user to navigate through quick looks in a GIS-like interface, with the possibility to add 
administrative boundaries, towns, streets, points of interest, etc.;
	–	allow the user to download his final product.

2.The direct readout system

Aim of this work is to integrate a web application with a direct readout system set up to acquire AVHRR 
images from NOAA-12, -15, -17 and -18 polar satellites passes. The direct readout system is a Seaspace [1] 
Terascan TL200 station which collects data from polar satellites; it is composed of a 45 cm dish  tracking 
antenna, a receiver, a bit synchronizer and a Linux acquisition control workstation. The workstation hosts 
Terascan software which makes passage acquisition scheduling, automatic pre-processing and image 
manipulation (visualization, integration, etc.) by means of TeraVision stand alone desktop package. Terascan  
works with a proprietary data format (Terascan Data Format - TDF) but  the post-acquisition processing chain 
can be configured to export the  processeded images in different formats, among others, for instance, the 
common format Hierarchical Data Format (HDF [2]).
In particular, Terascan was configured for: i) automatically acquiring all the passes of the satellite operative to 
date; ii) registering the images in a standard window of 1200 x 1200 km, which include Italy and part of the 
neighbour countries (central coordinate 42 1.29 N, longitudine 13 2.08 E); iii) projecting them in a UTM 
reference; and iv) exporting the final product in HDF. Any area of interest inside the satellite pass can be 
defined using a specific Terascan command (master) to be included in the pre-processing scripts. The 
acquired products are temporanery stored on the local file system in order to be successively downloaded by 
the web application.

3. The product classification system

Once acquired, the images are downloaded and classified on a workstation which represents the Web/DBMS/
Application server in charge to host the developed software modules. The first module of the application is in 
charge of: i) the ingestion of the remotely sensed products; ii) the extraction of a set  of standard information 
that will represent their metadata file; iii) the storage of these metadata in an ad-hoc database; and iv) moving 
the  datasets in an appropriate location on the local file system. The  developed database and, in particular, 
the metadata module, are designed  following the recommendations provided by ISO TC/211 standard [3],  
package 19115-2. The set of information chosen to represent the metadata file contains: product 
identification and extension, data quality, platform and mission, spectral properties, maintenance, spatial 
representation, reference system and other general information. The database management system (DBMS) 
used to implement the archive is PostgreSQL [4] with the  extension provided by PostGIS module [5]. PostGIS 
adds  a powerful support to PostgreSQL in dealing with geographic objects, allowing PostgreSQL to be used as 
a backend spatial database for geographic information systems. A "spatial" addition to the DBMS is needed as 
our system  must be able to manage data coming from different ground stations  (thus covering different 
areas of Earth surface), so that users can perform queries using, for instance, geographic criteria.  Moreover, 
as it will be discussed in the next section, the GIS enabling facility allows to store into the database other 
information, such as administrative boundaries, town locations, streets, etc.

4. The front end

The application interface includes several modules which allow users to search data in the archive, to 
immediately see  the current daily product, to display and navigate the image quick looks and, finally, to 
download  the selected ones. More in detail, the application was developed using PHP (configured to work with 
MapScript module [6]) and following a modular strategy and an object oriented pattern, in order to allow an  
easy extension of the prototype. In particular, the PHP MapScript  module is a PHP dynamically loadable 
module that makes MapServer's MapScript functions and classes available in a PHP environment. The map 
server used is the University of Minnesota Mapserver [7] and for  developing the navigation interface we 
started from GMap PHP/ MapScript demo application provided by the DM Solutions Group [8].
The quick look visualization and navigation interface is developed using the above mentioned components 
and allows the display of each AVHRR channel separately, and the dynamical addition of administrative 
boundaries, towns, streets, etc. Besides, other canonical operations like zoom in/out, pan, restore, etc., are 
allowed as well.

5. Conclusions and future developments.

This paper describes a preliminary experience in developing a web application prototype for remotely sensed 
data archiving and management, using PostgreSQL with PostGIS, PHP with PHP MapScript and other free 
software packages. The application works in conjunction with  a commercial direct readout system by 
extending its functions. This experience demonstrates the versatility of the mentioned Open Source 
components in building different applications for Earth Observation systems. The application was in the test 
phase for the last six months and has demonstrated to be stable and easily expandible. The planned future 
developments will regard:  i) a more complex infrastructure having the capability to manage  data coming 
from different remote sensing data sources; and ii) an automatic activation of the data processing chain, for 
monitoring critical thresholds and providing an alert in case of possible critical situations. The Model View 
Controller design pattern and Java technology will probably be adopted, in order to obtain a more efficient 
maintenance and scalability of the infrastructure. The developed prototype is available at the URL: http://
oracolo.le.isac.cnr.it/mpcf/RS.

6. References

[1] SeaSpace Corporation. [http://www.seaspace.com]
[2] Hierarchical Data Format. [http://hdf.ncsa.uiuc.edu]
[3] ISO/TC 211 Geographic information/Geomatics. [http:// www.isotc211.org]
[4] Postgres SQL. [http://www.postgresql.org]
[5] PostGIS. [http://www.postgis.org]
[6] MapScript. [http://www.maptools.org/php_mapscript]
[7] MapServer. [http://mapserver.gis.umn.edu]
[8] DM Solutions Group. [http://www.dmsolutions.ca]