(miri-mrs-example)= # MIRI Observations of Vesta Family This tutorial outlines the observation planning for *Vesta* family members using the MIRI instrument during JWST Cycle 6. Our main objective is to study silicate composition, which requires achieving a target Signal-to-Noise Ratio (SNR) of 300 around 10μm. As the 10μm feature is accessible with both MIRI Low Resolution Spectrometer (LRS) and Medium Resolution Spectrometer (MRS), we use both modes. We place the following constraints: · Targets must be Vesta family members with a known diameter. · The sample must span a range of diameters to encompass different sizes. · Required SNR is 300 around 10μm. ## Identifying potential targets We start by identifying all *Vesta* family members that have a known diameter, a prerequisite for accurately modelling their thermal emission. We use the ``rocks`` package to access the SsODNet/BFT asteroid database. ```python import rocks # Load table of asteroid properties (SsODNet/BFT) bft = rocks.load_bft() # Select candidate targets: Vesta family members with known diameter is_vestoid = bft["family.family_name"] == "Vesta" has_diameter = bft["diameter.value"].notnull() vestoids = bft[is_vestoid & has_diameter] ``` This initial selection yields 2041 Vesta family members with known diameters at the time of writing, all of which are potential targets. ## Defining MIRI configurations We configure the MIRI MRS and LRS modes using ``jayrock``, setting the optics, exposure times, and readout patterns for each. ### MRS Mode For the MRS, we set a standard 4-point dither (``nexp=4``). To target the 10μm silicate feature, we initially set the aperture to Channel 2 and the Long disperser. ```python import jayrock miri_mrs = jayrock.Instrument('MIRI', mode='MRS') miri_mrs.detector.nexp = 4 # 4-pt dither miri_mrs.detector.readout_pattern = 'fastr1' # recommended for MRS miri_mrs.aperture = "ch2" miri_mrs.disperser = "long" ``` ### LRS Mode We configure the LRS for SLIT mode and specify a 2-point dither (``nexp=2``). Since LRS has only one valid optical component, no further configuration is required. ```python miri_lrs = jayrock.Instrument('MIRI', mode='LRSSLIT') miri_lrs.detector.nexp = 2 # 2-pt dither miri_lrs.detector.readout_pattern = 'fastr1' # recommended for LRS ``` ## Determining observable diameters To understand the diameter range achievable with each MIRI mode, we select the largest Vesta family member within a set of representative diameter bins. ```python import pandas as pd # ------ # Get largest Vestoid within diameter range diameter_bins = [1, 2, 4, 8, 12, 16, 24, 32, 48, 64, 96, 128, 256, 512, 1024] vestoids["diameter_bin"] = pd.cut( vestoids["diameter.value"], bins=diameter_bins, right=True, include_lowest=True ) vestoids_largest_per_diameter_bin = vestoids.loc[ vestoids.groupby("diameter_bin", observed=True)["diameter.value"].idxmax() ] print(vestoids_largest_per_diameter_bin[["sso_name", "diameter.value", "diameter_bin"]]) ``` ```python sso_name diameter.value diameter_bin 2000 WE8 1.998 (0.999, 2.0] 2000 QA163 3.995 (2.0, 4.0] Kollaa 7.922 (4.0, 8.0] Koskenniemi 11.172 (8.0, 12.0] Mila 13.243 (12.0, 16.0] Robelmonte 23.834 (16.0, 24.0] Ausonia 93.000 (64.0, 96.0] Vesta 525.400 (512.0, 1024.0] ``` (4) Vesta is by far the largest family member, followed by (63) Ausonia at 93km diameter. No family members have sizes between 24 and 64km. ### Estimating exposure times We now iterate through this sample to estimate the minimum exposure time required to achieve an SNR between 300 and 400 at 10μm. Using an SNR range (instead of a single value) speeds up the binary search, as the process stops once any ``ngroup`` and ``nint`` combination is found. The resulting exposure times are thus appropriate upper limits. Once we have our target list, we will refine the exposure times further. For each target, we: 1. Compute its ephemeris for Cycle 6. 2. Identify the date of minimum thermal flux (``thermal_min``) during its visibility window, representing the most challenging observation scenario. 3. Estimate the necessary exposure settings (``ngroup`` and ``nint``) for both MRS and LRS modes. ```python # ------ # Estimate texp to reach SNR 300-400 for idx, asteroid in vestoids_largest_per_diameter_bin.iterrows(): # Define target and compute ephemeris for JWST Cycle 6 target = jayrock.Target(asteroid["sso_name"]) target.compute_ephemeris(cycle=6) target.print_ephemeris() # Add total length of visibility windows to dataframe vestoids.loc[idx, "N_days_observable"] = len(target.ephemeris) # Get date of minimum thermal flux during visibility window date_obs = target.get_date_obs(at="thermal_min") # Compute exposure times for SNR targets for inst in [miri_mrs, miri_lrs]: # Estimate ngroup and nint success = inst.set_snr_target([300, 400], target, date_obs) if not success: print(f" Could not estimate SNR, skipping...") continue # Record results vestoids.loc[idx, f"nint_{inst.mode}"] = inst.detector.nint vestoids.loc[idx, f"ngroup_{inst.mode}"] = inst.detector.ngroup vestoids.loc[idx, f"texp_{inst.mode}"] = inst.texp vestoids.loc[idx, f"snr_{inst.mode}"] = inst.estimated_snr ``` Example outputs of these calculations for three targets are shown below. :::{dropdown} The case of (2000) QA163 \(2000) QA163 is observable with both MIRI MRS and LRS. ```python (92804) 2000 QA163: Ephemeris from 2027-07-01 to 2028-06-30 ├── Window 1: 2028-01-19 -> 2028-03-13 │ ├── Duration 55 days │ ├── Vmag 21.03 -> 20.15 │ └── Thermal @ 15um 14.05 -> 26.62 mJy ├── Window 2: 2028-05-29 -> 2028-06-30 │ ├── Duration 33 days │ ├── Vmag 20.22 -> 20.81 │ └── Thermal @ 15um 25.43 -> 17.07 mJy └── errRA/errDec in arcsec: 0.006 / 0.002 # MIRI MRS SNR estimation INFO [jayrock] Searching for minimum ngroup|nint to reach SNR range 300.0-400.0 at 10.85μm INFO [jayrock] nint=1 | ngroup=5 -> SNR=3.3 | Texp=0.9min INFO [jayrock] nint=1 | ngroup=100 -> SNR=176.2 | Texp=18.5min INFO [jayrock] nint=2 | ngroup=100 -> SNR=247.7 | Texp=37.2min INFO [jayrock] nint=3 | ngroup=100 -> SNR=301.5 | Texp=55.9min INFO [jayrock] Done. Setting ngroup=100, nint=3. # MIRI LRS SNR estimation INFO [jayrock] Searching for minimum ngroup|nint to reach SNR range 300.0-400.0 at 10.70μm INFO [jayrock] nint=1 | ngroup=5 -> SNR=25.7 | Texp=0.5min INFO [jayrock] nint=1 | ngroup=100 -> SNR=426.4 | Texp=9.3min INFO [jayrock] nint=1 | ngroup=52 -> SNR=308.1 | Texp=4.8min INFO [jayrock] Done. Setting ngroup=52, nint=1. ``` ::: :::{dropdown} The case of (63) Ausonia \(63) Ausonia is observable with MIRI MRS but saturates in LRS mode. ```python # Ephemeris query results (63) Ausonia: Ephemeris from 2027-07-01 to 2028-06-30 ├── Window 1: 2027-12-22 -> 2028-02-16 │ ├── Duration 57 days │ ├── Vmag 11.93 -> 10.84 │ └── Thermal @ 15um 7667.04 -> 18441.00 mJy ├── Window 2: 2028-05-02 -> 2028-06-30 │ ├── Duration 60 days │ ├── Vmag 10.61 -> 11.48 │ └── Thermal @ 15um 23630.42 -> 12551.21 mJy └── errRA/errDec in arcsec: 0.011 / 0.007 # MIRI MRS SNR estimation INFO [jayrock] Searching for minimum ngroup|nint to reach SNR range 300.0-400.0 at 10.85μm INFO [jayrock] nint=1 | ngroup=5 -> SNR=596.3 | Texp=0.9min INFO [jayrock] Done. Setting ngroup=5, nint=1. # MIRI LRS SNR estimation INFO [jayrock] Searching for minimum ngroup|nint to reach SNR range 300.0-400.0 at 10.70μm WARNING [jayrock] Observation warnings: {'full_saturated': 'Full saturation:\n There are 879 pixels saturated at the end of the first group. These pixels cannot be recovered.', 'partial_saturated': 'Partial saturation:\n There are 335 pixels saturated at the end of a ramp. Partial ramps may still be used in some cases.'} WARNING [jayrock] nint=1 | ngroup=5 -> FULL SATURATION. ERROR [jayrock] Minimum nint/ngroup saturated. Stopping search. Providing parameter 'bounds' might avoid this. Could not estimate SNR, skipping... ``` ::: :::{dropdown} The case of (4) Vesta \(4) Vesta saturates even with the lowest possible ``ngroup`` and ``nint`` settings in both MIRI MRS and LRS. ```python # Ephemeris query results (4) Vesta: Ephemeris from 2027-07-01 to 2028-06-30 ├── Window 1: 2027-11-01 -> 2027-12-26 │ ├── Duration 56 days │ ├── Vmag 8.05 -> 7.08 │ └── Thermal @ 15um 215806.57 -> 464579.59 mJy ├── Window 2: 2028-03-10 -> 2028-05-06 │ ├── Duration 58 days │ ├── Vmag 6.99 -> 7.83 │ └── Thermal @ 15um 519366.32 -> 273551.43 mJy └── errRA/errDec in arcsec: 0.017 / 0.001 # MIRI MRS SNR estimation INFO [jayrock] Searching for minimum ngroup|nint to reach SNR range 300.0-400.0 at 10.85μm WARNING [jayrock] Observation warnings: {'ifu_full_saturated': 'Full saturation:\n In the current wavelength slice, there are 1 pixels saturated at the end of the first group. These pixels cannot be recovered.', 'ifu_partial_saturated': 'Partial saturation:\n In the current wavelength slice, there are 4 pixels saturated at the end of a ramp. Partial ramps may still be used in some cases.', 'cube_full': 'There are 6730 total fully saturated pixels in the data cube.', 'cube_partial': 'There are 1839 total partially saturated pixels in the data cube.'} WARNING [jayrock] nint=1 | ngroup=5 -> FULL SATURATION. ERROR [jayrock] Minimum nint/ngroup saturated. Stopping search. Providing parameter 'bounds' might avoid this. Could not estimate SNR, skipping... # MIRI LRS SNR estimation INFO [jayrock] Searching for minimum ngroup|nint to reach SNR range 300.0-400.0 at 10.70μm WARNING [jayrock] Observation warnings: {'full_saturated': 'Full saturation:\n There are 3602 pixels saturated at the end of the first group. These pixels cannot be recovered.', 'partial_saturated': 'Partial saturation:\n There are 1389 pixels saturated at the end of a ramp. Partial ramps may still be used in some cases.'} WARNING [jayrock] nint=1 | ngroup=5 -> FULL SATURATION. ERROR [jayrock] Minimum nint/ngroup saturated. Stopping search. Providing parameter 'bounds' might avoid this. ``` ::: The table below shows the exposure times (in seconds). A value of ``NaN`` indicates that the target saturates even at the minimum exposure settings. ```python sso_name diameter.value texp_mrs texp_lrsslit 2000 WE8 1.998 15684.52608 1115.56608 2000 QA163 3.995 3352.24832 288.60416 Kollaa 7.922 1110.01600 88.80128 Koskenniemi 11.172 577.20832 55.50080 Mila 13.243 577.20832 55.50080 Robelmonte 23.834 77.70112 NaN Ausonia 93.000 55.50080 NaN Vesta 525.400 NaN NaN ``` The exposure times confirm that MIRI MRS (IFU mode) is less sensitive than MIRI LRS (slit mode). MRS can observe larger, brighter targets without saturation, yet it requires considerably longer exposure times for the smaller targets compared to LRS. LRS saturates for targets larger than ∼24 km, and Vesta saturates in both modes. We can visualise this trade-off by plotting the exposure time against the diameter: ```python import matplotlib.pyplot as plt fig, ax = plt.subplots() ax.scatter( vestoids_largest_per_diameter_bin["diameter.value"], vestoids_largest_per_diameter_bin["texp_mrs"] / 60, label="MIRI MRS", s=50, ) ax.scatter( vestoids_largest_per_diameter_bin["diameter.value"], vestoids_largest_per_diameter_bin["texp_lrsslit"] / 60, label="MIRI LRS", s=50, marker="s" ) ax.set(xscale="log", yscale="log", xlabel="Diameter / km", ylabel="Exposure time / min") ax.legend() plt.show() ``` ```{figure} gfx/vesta_lrsmrs_dark.png :class: only-dark :align: center :figwidth: 100% **Fig. 1:** *Required total exposure times for SNRs between 300-400 versus diameter for selected Vestoids.* ``` ```{figure} gfx/vesta_lrsmrs.png :class: only-light :align: center :figwidth: 80% **Fig. 1:** *Required total exposure times for SNRs between 300-400 versus diameter for selected Vestoids.* ``` ## Defining final observing strategy Based on the exposure time analysis: · MIRI MRS is suitable for targets down to approximately 4km diameter to meet the SNR requirement within a reasonable exposure time. · MIRI LRS is necessary for the smallest targets, those <4km. For this example, we select the five targets ranging from 8km to 93km for observations in MRS mode and the two targets under 4km for observations in LRS mode. ### Configuring MRS mode For MRS, we must calculate the required exposure settings across all four channels and three dispersers. Channels 2 and 3 cover the key silicate features, so we set their target SNR to 300. Channel 4 is less sensitive and is set to SNR 100. For Channel 1, which has low throughput and is observed concurrently with Channel 2, we use the minimum exposure settings (``ngroup=5``, ``nint=1``). ```python TARGETS_MRS = ["Kollaa", "Koskenniemi", "Mila", "Robelmonte", "Ausonia"] SNR_targets = { 'ch2': {'short': 300, 'medium': 300, 'long': 300}, 'ch3': {'short': 300, 'medium': 300, 'long': 300}, 'ch4': {'short': 100, 'medium': 100, 'long': 100}, } ``` We now loop over our targets and instrument configurations to compute the required ``nint`` and ``ngroup`` for the APT. We always use 4-point dithers (``nexp=4``) for the target observations. We store these parameters in a dictionary to avoid recomputing them later. ```python MRS_OBSERVING_PARAMETERS = {} for target in TARGETS_MRS: # Define target and compute ephemeris for JWST Cycle 6 target = jayrock.Target(target) target.compute_ephemeris(cycle=6) # Get date of minimum thermal flux during visibility window date_obs = target.get_date_obs(at="thermal_min") for channel in ["ch1", "ch2", "ch3", "ch4"]: for disperser in ["short", "medium", "long"]: # Set aperture and disperser miri_mrs.aperture = channel miri_mrs.disperser = disperser if channel == "ch1": # Use minimum ngroup/nint miri_mrs.detector.ngroup = 5 miri_mrs.detector.nint = 1 else: # Get SNR target snr_target = SNR_targets[channel][disperser] # Estimate ngroup and nint based on SNR target miri_mrs.set_snr_target(snr_target, target, date_obs) print( f"{target.name:15s} | {channel:3s} | {disperser:6s} | " f"ngroup: {miri_mrs.detector.ngroup:2d} | nint: {miri_mrs.detector.nint:2d} | " f"texp: {miri_mrs.texp / 60:6.2f} min" ) if target.name not in MRS_OBSERVING_PARAMETERS: MRS_OBSERVING_PARAMETERS[target.name] = {} MRS_OBSERVING_PARAMETERS[target.name][(channel, disperser)] = { "ngroup": miri_mrs.detector.ngroup, "nint": miri_mrs.detector.nint, } ``` ```python MRS_OBSERVING_PARAMETERS = { "Kollaa": { ("ch1", "short"): {"ngroup": 5, "nint": 1}, ("ch1", "medium"): {"ngroup": 5, "nint": 1}, ("ch1", "long"): {"ngroup": 5, "nint": 1}, ("ch2", "short"): {"ngroup": 83, "nint": 4}, ("ch2", "medium"): {"ngroup": 69, "nint": 2}, ("ch2", "long"): {"ngroup": 74, "nint": 1}, ("ch3", "short"): {"ngroup": 62, "nint": 1}, ("ch3", "medium"): {"ngroup": 47, "nint": 1}, ("ch3", "long"): {"ngroup": 40, "nint": 1}, ("ch4", "short"): {"ngroup": 28, "nint": 1}, ("ch4", "medium"): {"ngroup": 48, "nint": 1}, ("ch4", "long"): {"ngroup": 91, "nint": 3}, }, "Koskenniemi": { ("ch1", "short"): {"ngroup": 5, "nint": 1}, ("ch1", "medium"): {"ngroup": 5, "nint": 1}, ("ch1", "long"): {"ngroup": 5, "nint": 1}, ("ch2", "short"): {"ngroup": 86, "nint": 2}, ("ch2", "medium"): {"ngroup": 72, "nint": 1}, # ... }, # ... } ``` #### Checking SNR and saturation We check the full SNR curve at both the date of minimum and maximum thermal flux to confirm the observation avoids saturation across the full visibility window. ```python for target, params in MRS_OBSERVING_PARAMETERS.items(): observations = [] print(f"Target: {target}") target = jayrock.Target(target) target.compute_ephemeris(cycle=6) date_obs_thermal_min = target.get_date_obs(at="thermal_min") date_obs_thermal_max = target.get_date_obs(at="thermal_max") for (channel, disperser), settings in params.items(): ngroup = settings["ngroup"] nint = settings["nint"] miri_mrs.aperture = channel miri_mrs.disperser = disperser miri_mrs.detector.ngroup = ngroup miri_mrs.detector.nint = nint obs_thermal_min = jayrock.observe(target, miri_mrs, date_obs_thermal_min) obs_thermal_max = jayrock.observe(target, miri_mrs, date_obs_thermal_max) observations.append(obs_thermal_min) observations.append(obs_thermal_max) # Save plot to file jayrock.plot_snr( observations, show=False, save_to=f"{target.name}_miri_mrs_snr.ng" ) ``` :::{dropdown} SNR of \(1929) Kollaa Below is the SNR plot for (1929) Kollaa at both the minimum and maximum thermal flux dates. All SNR targets are reached without saturation at both dates. Channel 1 will not yield much information, however, it is observed "for free" with channel 2. ```{figure} gfx/Kollaa_miri_mrs_snr_dark.png :class: only-dark :align: center :figwidth: 100% **Fig. 2:** *SNR of MIRI observations of (1929) Kollaa.* ``` ```{figure} gfx/Kollaa_miri_mrs_snr.png :class: only-light :align: center :figwidth: 100% **Fig. 2:** *SNR of MIRI observations of (1929) Kollaa.* ``` ::: :::{dropdown} SNR of \(63) Ausonia \(63) Ausonia is saturating in channels 2 and 3 at the dates of maximum thermal flux. This is shown in the console log (not shown here) and in the figure below, where only one SNR curve is plotted in these channels (SNR is ``NaN`` in case of saturation). We already use the minimum recommended ``ngroup``, ``nint``, and ``nexp`` values for this target. In this case, we could either go below (eg go towards a 2-point dither pattern) or restrict the observations to dates of low thermal flux only. This is left as exercise to the reader. ```{figure} gfx/Ausonia_miri_mrs_snr_dark.png :class: only-dark :align: center :figwidth: 100% **Fig. 3:** *SNR of MIRI observations of (63) Ausonia.* ``` ```{figure} gfx/Ausonia_miri_mrs_snr.png :class: only-light :align: center :figwidth: 100% **Fig. 3:** *SNR of MIRI observations of (63) Ausonia.* ``` ::: ### Configuring LRS mode We calculate the settings for the two smallest targets using MIRI LRS, aiming for an SNR of 300 at 10μm. ```python TARGETS_LRS = ["2000 WE8", "2000 QA163"] SNR_TARGET = 300 LRS_OBSERVING_PARAMETERS = {} for target in TARGETS_LRS: # Define target and compute ephemeris for JWST Cycle 6 target = jayrock.Target(target) target.compute_ephemeris(cycle=6) # Get date of minimum thermal flux during visibility window date_obs = target.get_date_obs(at="thermal_min") # Compute exposure times for SNR target at 10 micron miri_lrs.set_snr_target(SNR_TARGET, target, date_obs, wave=10) print( f"{target.name:15s} | " f"ngroup: {miri_lrs.detector.ngroup:2d} | nint: {miri_lrs.detector.nint:2d} | " f"texp: {miri_lrs.texp / 60:6.2f} min" ) LRS_OBSERVING_PARAMETERS[target.name] = { "ngroup": miri_lrs.detector.ngroup, "nint": miri_lrs.detector.nint, } ``` ```python LRS_OBSERVING_PARAMETERS = { "2000 WE8": {"ngroup": 65, "nint": 2}, "2000 QA163": {"ngroup": 36, "nint": 1}, } ``` #### Checking SNR and saturation We confirm that the LRS observations also avoid saturation at the date of highest thermal emission. ```python for target, params in LRS_OBSERVING_PARAMETERS.items(): observations = [] print(f"Target: {target}") target = jayrock.Target(target) target.compute_ephemeris(cycle=6) date_obs_thermal_min = target.get_date_obs(at="thermal_min") date_obs_thermal_max = target.get_date_obs(at="thermal_max") miri_lrs.detector.ngroup = params["ngroup"] miri_lrs.detector.nint = params["nint"] obs_thermal_min = jayrock.observe(target, miri_lrs, date_obs_thermal_min) obs_thermal_max = jayrock.observe(target, miri_lrs, date_obs_thermal_max) observations.append(obs_thermal_min) observations.append(obs_thermal_max) # Save plot to file jayrock.plot_snr( observations, show=False, save_to=f"{target.name}_miri_lrs_snr.png" ) ``` The results show no saturation issues for the two small LRS targets at either thermal extreme. :::{dropdown} SNR of \(2000) QA163 Bewlow is the SNR plot for (2000) QA163 at both the minimum and maximum thermal flux dates. The SNR target of 300 is reached without saturation at both dates. ```{figure} gfx/2000qa163_miri_lrs_snr.png :class: only-light :align: center :figwidth: 100% **Fig. 4:** *SNR curve of MIRI LRS observation of (92804) 2000 QA163.* ``` ```{figure} gfx/2000qa163_miri_lrs_snr_dark.png :class: only-dark :align: center :figwidth: 100% **Fig. 4:** *SNR curve of MIRI LRS observation of (92804) 2000 QA163.* ``` ::: ## Verification with JWST ETC and APT The ``MRS_OBSERVING_PARAMETERS`` and ``LRS_OBSERVING_PARAMETERS`` dictionaries contain all required information to set up the observations in APT. To verify that the SNRs and saturation levels match between Jayrock and the JWST ETC, we can export the spectra for each target and import them into the ETC. ```python TARGETS = TARGETS_MRS + TARGETS_LRS for target in TARGETS: target = jayrock.Target(target) target.compute_ephemeris(cycle=6) date_obs = target.get_date_obs(at="thermal_min") target.export_spectrum(f"{target.name}_spectrum_thermal_minimum.txt", date_obs=date_obs) ```