The Global Precipitation Measurement (GPM) mission will initiate a new era in precipitation measurement in terms of its global extent and frequency of sampling. Plans call for a GPM constellation consisting of eight spacecraft, with each spacecraft carrying a conical-scanning, microwave radiometer among its instrument complement.
One member of the constellation, designated the Core observatory, is at the heart of GPM, with assets and instrumentation contributed by NASA and the Japanese Aerospace Exploration Agency (JAXA). The Core observatory is uniquely instrumented with two crosstrack scanning radars, the Dual-frequency Precipitation Radars (DPRs), and a conical-scanning radiometer, the GPM Microwave Imager (GMI). This instrumentation enables the Core spacecraft to serve as both a precipitation standard and as a radiometric standard for the other GPM constellation members.
NASA is acquiring the GMI instruments through a commercial procurement. The contract for GMI was awarded in March 2005 to the Ball Aerospace & Technology Corp. of Boulder, CO. Launch of the Core observatory is planned for December 2010.
Conical Scanning Technology
The conical-scanning geometry of the GMI is illustrated in Figure 1. The off-nadir angle defining the cone swept out by the GMI is set at 48.5 degrees, which represents an earth-incidence angle of 52.8 degrees. Rotating at 32 rotations per minute, the GMI will gather microwave radiometric brightness measurements over a 140-degree sector centered about the spacecraft ground track vector. The remaining angular sector is used for performing calibration; i.e. observation of cold space as well as observation of a hot calibration target. The 140-degree GMI swath represents an arc of 1,170 km on the Earth surface. For comparison, the DPR instrument is characterized by cross-track swath widths of 245 km and 120 km, for the Ku and Ka-band radars, respectively. Only the central portions of the GMI swath will overlap the radar swaths (and with approximately 67-second duration between measurements due to the geometry and spacecraft motion). These measurements within the overlapped swaths are important for improving precipitation retrievals, and in particular, the radiometer-based retrievals.
The GMI is equipped with nine microwave channels. GMI channels lie within protected bands. GMI beam efficiencies for all channels will exceed 90% where beam efficiency is defined as the percentage of energy collected from an isotropic scene within the solid angle defined by 2.5 times the channel half-power beam widths and approximating the antenna main lobe between first nulls.
GMI Concept and Design
The Ball Aerospace concept for the GMI is illustrated in Figure 2. The GMI employs an offset parabolic antenna with an aperture size of 1.2 m. The antenna subsystem includes four feedhorns serving the nine channels. Each frequency is allocated an independent feedhorn with the exception of a shared feedhorn for the 18.7-GHz and 23.8-GHz channels. The antenna subsystem and receiver electronics rotate at 32 rotations per minute. A stationary thermal shroud, with an opening to cold space, surrounds the rotating instrument subsystems. The instrument will be responsible for its own momentum compensation. The control circuitry and logic governing instrument spinning and momentum compensation is contained within the instrument controller assembly. The instrument controller assembly and momentum wheel, providing momentum compensation, are mounted beneath the shelf supporting the GMI sensor.
Of particular interest are design features enabling superior instrument calibration. These design features include: (1) a four-point calibration technique employing an internal noise diode on each channel in conjunction with the standard cold sky and hot load calibration targets, (2) a low emissivity annular ring on the instrument deck for thermally isolating the hot load, and (3) a well-monitored hot calibration load with 14 platinum resistance thermometers placed within the target.
The four-point technique provides a polynomial fit to the instrument response that typically, in other spaceborne radiometers, is estimated by a linear fit. The conventional hot and cold targets are measured as per the typical total power calibration scenario. However, in addition, a noise diode is switched on momentarily during the hot load and cold sky target views, providing two additional temperature points available for calibration. The benefit of the technique is a more accurate instrument calibration.
The annular ring on the instrument deck is constructed of a metallic-coated material of low heat capacity and characterized by low emissivity at infrared and microwave wavelengths. The ring serves to radiatively isolate the hot load target from the instrument deck while the instrument deck rotates beneath it. In addition, the hot load and annular ring have enveloping shrouds to prevent solar heating either directly or through reflection. This design will help minimize temperature gradients within the hot load. All channels of the GMI will be calibrated to an accuracy of 1.35 K or better where this calibration value applies to the GMI main lobe temperature.
Spatial Resolution and Sampling
Figure 3 illustrates the instantaneous field of views (IFOVs) of the GMI channels and their translation from one scan to the next. The figure illustrates the relative size of the channel footprints, the common off-nadir angle for all channels, and the inter-scan distance resulting from the spin rate and spacecraft ground speed. (The figure does not reflect the instantaneous projections of the beams per the feedhorn GMI channel footprints in the Figure 3 layout.) For comparison between the GMI and DPR instruments, the IFOV for the DPR radars (identical for both radars) is illustrated. For channels 1 through 7, the IFOVs enable spatial contiguity in the cross-scan dimension. For channels 8 and 9, the IFOVs satisfy a minimum coverage in the cross-scan dimension of 50%.
The 1.2-m-diameter aperture of GMI provides excellent spatial resolution for channels 1 through 5, the channels for which the entire aperture is utilized in beam formation. These GMI channels offer fine spatial resolution when compared to other conical-scanning radiometers.
The choice of sampling times for the GMI is governed by the desire to achieve "Nyquist" spatial sampling in the along-scan dimension of the swath. In addition, samples from individual channels must be co-registered on the Earth surface. Sample times are slightly larger than integration times due to latencies inherent to the digital sampling electronics. To satisfy the Nyquist criterion, all channels will be sampled at a minimum of two times as the GMI scans through a single IFOV. To guarantee co-registration on the Earth, sample times for each channel will be integral multiples of each other. Electronic time delays between channel sampling will account for the fixed angular differences in along-scan beam pointing due to the multiple feedhorn design.
The GMI radiometer will serve as a transfer standard in two contexts: as a radiometric transfer standard for the other radiometers of the GPM constellation, and as a precipitation transfer standard for the retrievals of the GPM constellation. Both transfer standards represent areas of scientific research.
In the first context of the radiometric transfer standard, the GMI radiometric calibration will serve as a reference for other radiometers. In this method, the brightness temperature calibration of constellation member radiometers will be adjusted to achieve a common basis with that of the GMI. This technique will reduce precipitation retrieval differences between sensors due to biases from inter-sensor calibration. Referencing calibration to the GMI requires that the GMI maintain excellent calibration and calibration stability through its life. In operation, the ability to create a common calibration from the GMI will depend upon statistical data from spacecraft intersection events, i.e. the viewing of common earth scenes. Such intersection comparisons are complicated by factors such as intervening clouds and precipitation, different look angles, different footprint sizes, and time delays and sample spatial separations specific to each intersection.
The second context refers to a precipitation transfer standard. Specifically, this concerns the measurement synergy created by the GMI and the Dual-frequency Precipitation Radars aboard the Core observatory. The mutual overlap of actively sensed, vertically profiled, radar data at two frequencies in combination with the multi-channel passive data is a unique capability of the Core observatory. The GPM radars will be used to accurately measure, via reflectivity and estimates of attenuation, the vertical profiles of the clouds and precipitation, including the drop size distribution. This radar profiled data will be compared to the GMI radiometric retrieval and its derived profile, which are presumed less accurate than that from the radars. As part of the precipitation transfer standard, the radar-measured profiles, and associated radiometer brightness temperatures, are included in an improved database available to the GMI as well as other GPM radiometers.
GPM will transfer Core observatory GMI data to the user community at very short latency. In particular, data from GMI will be continuously transferred via a Tracking and Data Relay Satellite System (TDRSS) Multiple Access (MA) link through White Sands, NM, and into the availability of the user community, typically 15 minutes or less from the time of measurement.
This article was written by S.W. Bidwell, G.M. Flaming, J.F. Durning, and E.A. Smith of NASA's Goddard Space Flight Center in Greenbelt, MD. For more information, click here .