IEEE Geoscience and Remote Sensing Magazine - June 2014 - 69

country is free to modify the
ITU table for services within
its borders [1].
Each allocated service is generally granted either primary
or secondary status. Secondary
services shall not cause harmful
interference to primary services.
They also cannot claim protection from harmful interference
caused by a primary service;
however they can claim protection from harmful interference
from sources of the same or other secondary services. Current
Radio Regulations have allocated frequencies between 9 kHz
and 275 GHz. It is important
for GRSS FARS to represent the
spectrum needs of its community at a spectrum management
meeting such as WRC.

taBLe 1. rfi-tHreatened passive eess frequency aLLOcatiOns.
Bands (GHz)

scientific OBservatiOns spaceBOrne instruments rfi LeveL and sOurces

1.37-1.40*
1.40-1.427

Soil moisture, sea surface
Salinity, sea surface wind,
vegetation index

SMOS,
Aquarius, SMAP

High; out of band emissions mostly from air
surveillance radars.

6.425-7.25**

Soil moisture, sea surface
temperature, precipitation

AMSR-2, WindSat

Moderate (especially
over the U.S.A.)

10.6-10.7

Precipitation, cloud liquid
water, sea surface wind
speed, sea surface temperature

TMI, AMSR-2, WindSat
GPM GMI

Moderate (especially
over Europe)

18.6-18.8

Precipitation, cloud liquid
water, snow cover, sea surface wind speed, sea ice

JASON-2 AMR, AMSR-2,
WindSat

Moderate; potentially
from satellite TV service
signals.

22.21-22.5
23.6-24

Atmospheric water vapor,
Sea surface wind speed,
sea ice, precipitation, snow
cover

SSM/I, JASON-2 AMR, TMI,
AMSR-2, GMI, AMSU-A

Moderate; vehicle anticollision radars

31.3-31.8
36-37

Precipitation, cloud liquid
water, snow cover, sea surface wind speed, sea ice

AMSU-A, SSM/I, TMI, AMSR-2
JASON-2 AMR, WindSat,
SSM/IS, GPM GMI

Low; new sources observed off oil platforms
near the Indian subcontinent

50.2-50.4
51.4-59.3

Atmospheric temperature
profiling

TOVS (MSU), SSM/T,
AMSU-A, SSM/IS

Moderate: potential for
RFI due to spectrumsharing rules at 55-57

*Secondary allocation
2.1 Frequency AllocAtions
**Limited protection
And rAdio Frequency
interFerence in PAssive
remote sensing
Though frequency allocations exist for microwave remote
sensing, RFI is still a significant issue. Table 1 shows a list
of the most relevant frequency bands allocated to passive
remote sensing EESS (Earth Exploration Satellite Service)
[2],[3] with their main scientific uses and their spaceborne sensors which are currently-or are scheduled to
begin-operating in that frequency band. The expected
level of RFI and its sources are also indicated in Table 1.
Except for the secondary 1.37-1.40 GHz allocation and
the 6.425-7.25 GHz limited-protection band, all allocations are primary.
When the band is shared with other services, RFI
degrades the quality of the remote sensing data. Passive
sensors are especially vulnerable to RFI because of the
very low levels of the detected signal that carries information. The last column in Table 1 reports level and likely
sources of observed RFI. Even in protected and nonshared bands, RFI can still originate from sources emitting in adjacent bands. As noted in Table 1 and the figures
below, even though many of the allocations are primary
or protected, RFI still plagues the incoming brightness
temperature (T b) signals. Frequency allocations from
82 to 275 GHz were not reported, since the current RFI
threat is low for those channels.
An example of the RFI threats stated in Table 1 is
shown in Fig. 1 which indicates the RFI observed in three
of the most used bands in passive microwave remote
sensing (1.4 GHz, 6 GHz and 10 GHz). Due to different spectrum allocation rules in different ITU regions,

june 2014

ieee Geoscience and remote sensing magazine

the geographical distribution of RFI varies depending on
the frequency. Data acquired over the United States of
America show corruption in the 6 GHz region, whereas
the United Kingdom and Italy are more affected at 10
GHz. Fig. 1 shows RFI hot spots observed by AMSRE in December 2010 and Aquarius in December 2011,
obtained by running a peak hold algorithm over approximately a month of data.
Fig. 1 is a snapshot of the current occurrence of RFI
at the respective frequencies and does not represent the
increase in RFI at various bands over time. There have
been studies indicating a clear increasing RFI trend over
the last decade. A study done by McKague et al. [4] gives
an example of increasing occurrence of RFI at K-band
from 2005 to 2010. Fig. 2 illustrates the temporal evolution of the spectral difference between the brightness
temperature T b of the AMSR-E 18.7 GHz and 23.8 GHz
channels. A negative value of the T b difference represents
RFI in the 18.7 GHz channel. The bottom plot (for January 2009) shows an "arc" of RFI observed over the northern regions of United States. This broad arc is attributed
by McKague et al. [4] to reflections off a Direct TV satellite signal, based on the relative position and incidence
angles of the satellites.
Given the presence of interference at bands of interest
to the remote sensing community and the vulnerability of
passive microwave systems to man-made sources, over the
past decade a significant amount of work has been done to
develop RFI detection and mitigation algorithms.
69

Table of Contents for the Digital Edition of IEEE Geoscience and Remote Sensing Magazine - June 2014