Chang'e MRM Derived Data (MAP)
MAP – Chang'e-1 & Chang'e-2 Time-Binned Microwave Radiometer Brightness Temperature Maps
| Data Notice
The derived map products include three types of map-projected brightness temperature (TB) products generated from CE-1 and CE-2 MRM calibrated data, which contain known calibration issues, including orbit-to-orbit temperature offsets and elevated noise. The "temp" maps are direct deconvolution outputs and most directly inherit these calibration artifacts. The "datminus" maps (observed minus physically modeled TB) and "latshift" maps (observed minus latitudinally fitted TB) are designed to reduce sensitivity to absolute calibration offsets and are generally better suited for detecting relative spatial variations in surface properties. Although bundle creators have excluded some observations with identified errors, users should be cautious with these derived products. For full details on derivation methods and calibration considerations, see the Bundle README and Primary Bundle Documentation. |
Instrument: Microwave Radiometer (MRM)
PDS4 Bundle:urn:nasa:pds:chang_e_microwave_processedDOI: 10.17189/9nfs-ed98
For more information about MRM derived products, see the Primary Documentation for chang_e_microwave_processed.
The bundle includes 4 types of map-projected products: "temp", "latshift", "datminus", and "tbmod". Each orbiter/channel combination has three associated map products: a “temp” product, a “latshift” product, and a “datminus” product. Each channel also has an associated “tbmod” product. (Note that the "tbmod" maps are in the bundle's miscellaneous collection, as they do not use CE MRM observational data as an input. The others are in the bundle's data collection. )
Temp (brightness temperatures) products
Unlike the other map product types, "temp" products contain two distinct map types. TEMP maps contain lunar brightness temperatures derived by deconvolving MRM antenna temperatures. STDEV maps contain biased estimates of weighted standard deviation for the antenna temperature values used to derive the values in their corresponding TEMP maps (these values may be useful for both physical interpretation and data quality assessment).
Each "temp" product contains either 19 (CE-1) or 27 (CE-2) HDUs:
* PRIMARY (FITS stub)
* one TEMP HDU for each 2-hour time bin with valid data (8 or 12 HDUs total)
* one STDEV HDU for each 2-hour time-bin with valid data (8 or 12 HDUs total)
* LATITUDE
* LONGITUDE
Datminus products
The DATMINUS maps in these products contain model-subtracted derived brightness temperatures. Each is created by subtracting the corresponding TBMOD map of its matching “tbmod” product from the corresponding TEMP map of its matching “temp” product (after cropping that TEMP map to the latitude range of the TBMOD map). Their values can be interpreted as differences between measured temperatures at each grid cell and expected temperatures at each grid cell, where this expectation is set by a physical model that takes into account channel frequency, local time, latitude, thermal conductivity gradient, and albedo. Because this model does not account for dielectric loss, much of the regional and diurnal variability in these maps (especially at channels 1 and 2) is generally dependent on concentrations of the titanium-containing mineral ilmenite.
Each "datminus" product contains either 11 (CE-1) or 15 (CE-2) HDUs:
* PRIMARY (stub)
* one DATMINUS HDU for each 2-hour time bin with valid data (8 or 12 HDUs total)
* LATITUDE
* LONGITUDE
Latshift products
The LATSHIFT maps in these products contain latitudinally adjusted derived brightness temperatures. Each is created by fitting a simple latitudinal model of temperature (`a + cos(lat) ** b`, with `a` and `b` as free variables) to the corresponding TEMP map of its matching “temp” product, then subtracting the fitted curve from the TEMP map. Their values are differences between measured temperatures at each grid cell and expected temperatures at each grid cell, where this expectation is set *only* by the latitudinal trend in data taken by this particular orbiter in this particular channel during this particular range of local times and incorporated into this particular TEMP map.
These maps were provided in addition to the DATMINUS maps because there are many confounders in these data, including simple sample selection bias produced by offsets between different orbits, that make it difficult to model the joint dependence of temperature on LTST and latitude. By restricting the population of samples in the model, the LATSHIFT maps provide a stricter, although *strictly relative*, view of temperature variation. This means that some features may be apparent in these maps that are not apparent in the DATMINUS maps.
Each "latshift" product contains either 11 (CE-1) or 15 (CE-2) HDUs:
* PRIMARY
* a LATSHIFT HDU for each 2-hour time bin with valid data (8 or 12 HDUs total)
* LATITUDE
* LONGITUDE
Tbmod products
The TBMOD maps in these products contain modeled brightness temperatures derived by interpolating Kaguya SELENE and LRO Diviner, LOLA and LROC data within the parameter space of a physical heatflow model. This process is described in more detail in the “Thermal Brightness Modeling” section in the Primary Documentation for chang_e_microwave_processed. Because they do not use CE MRM data as an input, there is only one “tbmod” product for each channel.
Each “tbmod” product contains 15 HDUs:
* PRIMARY
* one TBMOD HDU for each 2-hour time bin (12 HDUs total)
* LATITUDE
* LONGITUDE
Relationships Between Derived Map Products
“temp” products are direct outputs of the deconvolution pipeline. They are derived from the time-series values in the concatenated data table of their corresponding orbiter. “latshift” products are derived from the “temp” product of the same orbiter and channel. “datminus” products are derived from the “temp” product of the same orbiter and channel and the “tbmod” product of the same channel.
Although the source products for “temp”, “latshift”, and “datminus” FITS files are explicitly stated in their PDS4 labels, they can also be determined simply from their filenames:
* `[orbiter]_[channel]_temp.fits` is derived from `[orbiter]_mrm.fits`.
* `[orbiter]_[channel]_latshift.fits` is derived from `[orbiter]_[channel]_temp.fits`.
* `[orbiter]_[channel]_datminus.fits` is derived from `[orbiter]_[channel]_temp.fits` and `[channel]_tbmod.fits`.
MRM derived map products have the following file names:
Filenames of "temp", "latshift", and "datminus" products follow the pattern:
[orbiter]_[channel]_[product type]_32ppd.[extension]`
where:
orbiter = "ce1" or "ce2". It indicates which orbiter's MRM data were used to produce the product.
channel = "t1", "t2", "t3", or "t4", corresponding respectively to the MRM channels centered at 3, 7.8, 19.35, and 37 GHz. Literally, it is the name of the column that the deconvolution program(`run_deconv`) loaded from the corresponding binary source data table ([orbiter]_mrm.fits) described in the "Calibrated Data" section.
product type = "temp", "latshift", or "datminus".
extension = "xml" (the product's PDS4 label) or "fits" (the product's FITS file).
Because "tbmod" products are derived from non-MRM observational data, their filenames do not include an orbiter code, and instead follow the pattern:
[channel]_tbmod_32ppd
where:
"channel" has the same basic meaning as in the MRM-derived map products. Literally, it indicates the frequency value given to the brightness temperature modeling program (`map_brightness`).
*Note: the "32ppd" in these filenames does not imply that this bundle contains similar map products at other resolutions. This field was included partly for consistency with certain model inputs and partly because data providers may, in the future, choose to produce map products at other resolutions.*
In ODE, MRM derived products have the following product IDs:
ce1_[channel]_[product type]_32ppd
ce2_[channel]_[product type]_32ppd
[channel]_tbmod_32ppd