IEEE
Geoscience and Remote Sensing Society

Abbreviation: GRSS, S Code 29

 The IFT Committee has identified the following key technology areas which pose challenge to the remote sensing community in the coming years.

1. Reducing size, power requirements, and cost of remote sensing instrumentation to provide more feasible access to space. Applicable technologies include:

• Deployable large aperture antennas
• MMIC technologies for millimeter- and submillimeter-wave components
• Low noise receivers (superconducting technologies, cryogenic coolers)
• Digital signal processors, high speed correlators, and fast (> 1 Gsample/sec) 16-bit Analog to digital converters for real time data systems.

2. Developing infrastructure in the rapidly changing world of computing and networking to process and disseminate the terabytes of data per day anticipated from remote sensing platforms.

• Efficient search and retrieval of archived data
• Data compression
• Improved utilization of the internet for data distribution
• Standardization of remote sensing data formats

3. Developing algorithms which fuse different data sources to derive physical parameters needed by the data user community.

• Standardization of instrument and algorithm calibration
• Desktop supercomputing power to incorporate remote sensing data into global climate models
• Cross calibration of remote sensing data with respect to instrument, spatial resolution, and time
• Distributed network computing for the operational assimilation of remote sensing data

Committee members' comments:

Curt Davis:

On the list you prepared I believe that reducing size, weight, and

and especially COST is one of the most important technology issues

the GRS faces. The other two issues are important and overlap a

lot with the interests of the Data Standards and Distribution

Technical Committee.

One issue that could be added to the list on #1 would be

the development of commercial applications and products

of remote sensing data to help subsidize the cost of satellite

missions. Some small efforts have been made in this area but

maybe the GRS AdCom should take a good look at this issue

to see what potential is there?

 *~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*

Dan Ziskin:

provide input to the formation of data policies

Umberto Spagnolini:

I agree with your list and I would suggest the following challenges closely

related to (but not necessarily only to) the GPR technology:

- Spread the access to remote-sensing data to wider community (e.g.,

broadband internet access);

- Data compression (e.g., handle archived data at different scales or with

different information contents);

- Include more detailed (not highly simplified) physical models into

algorithms used to derive the physical parameters;

- Reduce costs by developing algorithms for automatic processing of large

amount of data (or information) with reduced human interaction.

David Long:

How about "new system concepts for radars to make them smaller and lighter" as

well as improved hardware. For example MMIC's can be used to reduce the

size/weight/power of microwave sensors too.

*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*

C. Lin:

1. Reducing size, power reqs., and cost ...

All mentined areas seem reasonable, but it is difficult to put priorities on

the different items. As there are several bottle-necks in a given remote

sensing system, solving one of them immediately means hitting another one.

Deployable large aperture antennas only apply to space based systems whereas

MMICs and digital processors are applicable to all kind of remote sensing

systems.

One important parameter which one usually accepts as it is in any system

design is the receiver noise. It immediately imposes several system parameters

such as antenna gain, transmit power, signal integration length, spatial

resolution, etc... Imagine that we could one day reduce the receiver noise

by a factor of 2 or 4, or even more! This would open a completely new system

optimization possibilities (e.g. smaller antenna, lower transmit power,

shorter signal integration, higher spatial resolution, etc...).

Now the question is: "how could we drastically reduce the receiver noise?"

The obvious answer is an efficient receiver cooling. Whatever techniques

which could be developed to achieve an efficient and reliable cooling of

key receiver components/subsystems would be a significant step for generally

improving a large class of microwave instruments. Could the latest high

temperature superconductor technology contribute significantly to such a

reduction of receiver noise? Well, those are rather open questions.

A second area of interest could be the development of fault/failure tolerant

components and subsystems. Current engineering practice uses dedundancies

and vigorous quality assurances to reduce the probability of failure.

Still, there are countless examples of lost or delayed system deployments due

to critical single point failures (e.g. stucked antenna of Galileo, failed

navigation system of Ariane 501, stucked scanning mirror, unstable whisker

contacted mixer, etc...). Those critical components/subsystems should be

designed in the future in such a way that they are by themselves fault/failure

tolerant, without necessarily calling for an internal redundancy. This might

sound like a fairy-tale. But it might worth giving a thought.

2. Developing infrastructure ... + 3. Developing algorithms ...

Since remote sensing is an information technology, I would rather put those

two challenges into one. Your list of items looks rather complete. But I

would like to add the integration of local high resolution models (e.g.

hydrological model, coastal tide and current model, urban micro-climate model,

regional weather model, etc...) in addition to the global climate models.

Indeed, the climate community is a small one, as compared to the rest of the

world which rather needs more detailed information on regional or local

scales. One could expect in the future that those regional/local high

resolution models will be operational into which remote sensing data are

permanently assimilated. There will be no more clear distinction between the

actually measured data and model derived data. The continuous availability

of model derived information, rather than point-wise measured data would allow

users to retrieve information on an arbitrary location at an arbitrary point

in time. Therefore, direct accesses to those model derived information would

be necessary. I remember the time when I was constantly listening to the local

weather radio band of the San Francisco bay area. From the noon on in Summer,

I was in search of good wind for windsurfing. The wind/wave and tide

information was updated every two hours and they were available only at the

locations of buoys. Extrapolating those point-wise information to my favorate

sailing locations was a rather tricky buisiness, and I had to wait for further

two hours if the wind condition was bad everywhere. If a continuous update

of the information was made through running of a regional weather/current/

wave/tidal model, I could have retrieved the most up-to-date (best prediction)

information at any time at any of my favorate sailing spots by just connecting

to say Internet. I could have even known in advance how long the wind will

blow in that afternoon.

Well, I got somehow carried away ...

It is difficult to list exactly what technology items are relevant for such a

scenario to be possible. But your list of technology areas seem reasonable.

*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*

Reto Peter:

Somewhere in your list progress towards very low noise receivers should be included.

This concerns mainly the development of super-conducting (SIS) devices and

receivers in the mm- and particularly in the sub-mm range (airborne and space-borne

systems).

So the keywords would be : Low noise receivers using super-conducting SIS

technology

*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*

Sandy Weinreb:

I just returned from the SPIE Aerosense Conference in Orlando on the subject of

Passive Millimeter Wave sensors - This was about the sensors and phenonology for

the applicaitons of all-weather landing systems, military target location, and

contraband weapons search.

In the area of spaceborne passive millimeter wave sensors, I would make the

following comments:

1) MMIC radiometers at 54, 118, and 183 GHz are being developed under the NASA

IMAS program for nadir sounding of temperature and humidity. The radiometers

are being developed under contracts to TRW and Lockheed Sanders and have

specifications of mass < 2 Kg, power < 5 W, and sensitivity of 0.7K at 20 ms

integration time. A contact at JPL is Todd Gaier at 818-354-4402.

2) The IMAS program is for single-pixel sensors. A much greater advance can be

made by the development of MMIC millimeter wave focal plane arrays which measure

many points in one time interval and can eliminate the need for mechanical

scanning. A focal plane array with 75 pixels in the 100-140 GHz and 170-210

GHz bands was proposed as AMLS (Array Microwave Limb Sounder) to the NASA New

Millenium program. It received high marks but has not been funded at present.

A contact at JPL is Joe Waters at 818-354-3025.

3) Another area needing development is low-cost sensors for the sub-millimeter

range (200 GHz to 2500 GHz). Diode-based integrated circuits (MMIC's) and

micromachined antennas could replace the labor intensive and difficult assembly

of the discrete-diode, precision machined, receivers which are presently in

use.

John P. Kerekes:

Your list of challenges is very good and pretty much complete. I would

add one more aspect to the reducing size, weight, power, etc. and that

is efficient representation of the information in the data. One aspect

of this is data compression, but it also includes application-specific

metrics and the ability to retain all necessary scientific value in the

data. Particularly for EO systems like rapid scanning advanced GOES Imagers,

AIRS, or hyperspectral imagers the data rates are tremendous and will

overwhelm existing infrastructures, but we know significant redundancy exists

in the data; we just don't know how to exploit that in all situations.

I assume this comment is too late for your reply, but I'm sending it

anyway to let you know your messages do get read and there is still

interest out there.