GLOBAL DATA SETS FOR LAND APPLICATIONS FROM THE ADVANCED VERY HIGH-RESOLUTION RADIOMETER - AN INTRODUCTION

The Advanced Very High Resolution Radiometer (AVHRR) has become one of the most important sensors for monitoring the terrestrial environment at resolutions of 1 km to very coarse resolutions of 15 km and greate r. To make these data suitable for scientific and other applications c onsiderable effort has been devoted to the creation of global data set s. Experience has demonstrated that even for a relatively simple senso r such as the AVHRR, the task of creating global data set is fraught w ith difficulties and that a number of iterations have been necessary d espite considerable efforts in the specification of users' requirement s. Four types of data processing streams, overlapping in time, have oc curred in the creation of global data sets from the AVHRR. The first t hree data processing streams were all based on the reduced resolution, Global Area Coverage (GAC) data set, which is collected globally ever y day. In the first data processing stream a much reduced data set was created in the form of the Global Vegetation Index (GVI) product: rev ised improved versions of the product have been produced. In the secon d data processing stream an improved product was created by workers at NASA's Goddard Space Flight Center with higher spatial resolution but which until recently has only been available by continent. This has r esulted in the creation of a number of regional data sets. In the thir d data processing stream operational creation of global data sets at m oderately coarse resolution (c.8 km) is being initiated. The most nota ble example of this data processing stream is part of NASA's Pathfinde r project and stems in large part directly from the second data proces sing stream: it will involved production of a reprocessed improved glo bal data set for the period from 1982 to the present. In the fourth da ta processing stream the full potential of the AVHRR in terms of its s patial resolution is being realized, through the generation of a globa l data set at 1.1 km resolution data.