Refer to PSP 29th Perihelion Campaign page.
CONSENSUS PREDICTION (CSV, PDF table of coordinates)
The predicted footpoints were kindly provided by the PSP 29th Perihelion modeling team.
Encounter 29 Prediction Update 2/4: 2026/09/02
------------------------------------------------
This is the second of four footpoint predictions issued for Parker Solar Probe Encounter 29, the missions eighth to reach a perihelion distance of 9.86 Rs (which will be reached on Sept 4, 14:53 UT). The perihelion location and most of the encounter will be on the far side of the Sun, however the inbound phase yesterday and today down to 15 Rs will be connected on the west limb, and the outbound phase above 20 solar radii will be connected near the east limb. Therefore predictions are being issued around these two events as a pair yesterday and today (Sept 1 & 2, West limb targets), and then again on Sunday Sept 6 and Monday Sept 7 (East limb targets).
Magnetic Connectivity
---------------------
The faint coronal hole EUV feature from yesterday is still just visible at the west limb, between ARs 14518 amd 14515. We predict that Parker's instantaneous connection to this source will persist until approximately 2026/09/02 1900UTC and as the most recent connection prior to perihelion, may remain an interesting target for observers tomorrow.
After west limb passage, while the footpoints are on the far side, the consensus source remains clear with all footpoints targeting a positive polarity coronal hole which was visible on disk, extending over a broad range of southern latitudes, and was at disk center about 1 week ago.
Looking ahead to East limb passage, the footpoints are also largely converged to a negative polarity near equatorial coronal hole at mid-southern latitudes. This source was seen at disk center on August 16. Predictions suggest the footpoints will co-rotate on disk with this source starting from Sunday (9/6) at 19:00 UT.
Flare Likelihood (CCMC Flare Scoreboard)
----------------------------------------
As of 2026/09/02 at 12:00 UT, CCMC / ISWA Flare Scoreboard reported 24-hour average cumulative probabilities amounting to 82%, 12% and 1% for GOES C and above, M and above and X-class flares, respectively, slightly elevated since yesterday’s update. The strongest flare reported over the past 24 hours was a GOES C5.7 event peaking at 2026/090/01 at 21:10 UT and coming from the eastern limb (N12E88). NOAA active regions visible on the disk continue to include ARs 14515 and 14518 — 14523. Of them, only NOAA ARs 14521 and 14523 are located at eastern longitudes while all the rest are in the western solar hemisphere. Only NOAA ARs 14515, 14520 are located at southern latitudes, with every other active region in the north. There are no major changes in the corona since yesterday, with the reported coronal hole moving closer to the central meridian. Western longitudes where Parker footpoints are projected continue to be covered mostly by closed and active region magnetic fields.
From the CCMC CME Scoreboard, there does not seem to be an active Earth-directed CME at this time.
------------------
*** Please note that the "arrival time" and "emission time" and associated Tx/Ty coordinates for both are reported in the consensus CSV file. The attached plots show the "arrival time" (location of source at time that plasma will arrive at PSP). See Parker Footpoint Predictions - Meaning and Terminology for some discussion on these***.
Date and arrival time of plasma parcel at PSP, consensus carrington longitude (deg), latitude (deg), error in longitude, error in latitude, on-disk position of predicted source in X and Y (arcseconds) at arrival time, date and emission time of plasma parcel at the source, on-disk position of predicted source at time parcel is emitted. Each row is the updated source location each hour.
The consensus is generated by forming a distribution of footpoint predictions from all modelers for each hour period, and attempting to fit a Kent distribution. If the fitting fails, the median in longitude and latitude are quoted. If the fitting is successful, the quoted errors are formed by drawing random samples from the fitted distribution and computing the standard deviation in longitude and latitude of those samples. If the fitting fails, the quoted errors are the standard deviation in the longitude and latitude from the raw distribution of predictions. The full shape of the distribution is described by black contours in the associated plots on this website. More details about the procedure can be found at the following preprint of Badman et al. (2023) "Prediction and Verification of Parker Solar Probe Solar Wind Sources at 13.3Rs"
Please note the carrington coordinates (lon,lat) are valid from the quoted timestamp (in UTC) until the next timestamp. The helioprojective coordinates quoted (HP-Tx, HP-Ty) are computed from the carrington coordinate at the quoted timestamp (e.g. midnight UTC each day) and so are valid instantaneously at this time but will corotate with the Sun until the next quoted timestamp. For a discussion of the subtle difference in emission and arrival time and why both are included please see the slide deck
Individual model prediction tables of coordinates may be found in a Public DropBox. Files in the Public DropBox have three-letter identifiers indicating the associated model (see below).
Three-letter designation for Public DropBox: UCB. Kindly provided by Sam Badman. The model is a simple ballistic propagation from PSP down to the source surface assuming slow wind 360km/s, and then tracing this sub-PSP trajectory through a PFSS model to get footpoints at the photosphere. The source surface height here is 2.5Rs. The PFSS model is generated using various ADAPT maps with GONG and HMI as input, and the model is run using the open source pfsspy package. A more detailed explanation of the model and comparison to PSP E1 results are given here.
Three-letter designation for Public DropBox: PSI. Kindly provided by Pete Riley. For these predictions, PSI is using a combination of modeling approaches, including PFSS solutions, empirically-based polytropic MHD solutions, and a more sophisticated approach that includes the effects of waves and turbulence to heat the corona and the WKB approximation for wave pressures to accelerate the solar wind. Additionally, boundary conditions are derived from both HMI and ADAPT synoptic magnetograms. Together, these allow us to generate a rich set of ensemble realizations from which to make our optimal prediction, as well as pool them with other teams’ forecasts to derive a hyper-ensemble prediction.
Three-letter designation for Public DropBox: wsa. Kindly provided by Shaela Jones. The Wang-Sheeley-Arge (WSA) model is a combined empirical and physics-based model of the corona and solar wind. The coronal portion of the Wang-Sheeley-Arge (WSA) model is comprised of the Potential Field Source Surface (PFSS) and Schatten Current Sheet (SCS) models, where the output of the PFSS model serves as input to the SCS model. The solar wind portion of WSA consists of a simple 1-D kinematic propagation code that takes stream interactions into account in an ad-hoc fashion. It provides predictions of the solar wind speed and interplanetary magnetic field IMF polarity at any specified point in the inner heliosphere. The WSA model can use global maps of the photospheric magnetic flux measurements from a number of sources as its inner boundary condition; here we are using an ensemble of maps from the Air Force Data Assimilative Photospheric Flux Transport (ADAPT) model, based on input GONG magnetograms.
UAH predictions come from the University of Alabama, Huntsville Multiscale Fluid-Kinetic Simulation Suite (MS-FLUKSS, Pogorelov et al. (2014); Pogorelov (2023); Singh et al. (2022)), which can solve the Reynolds-averaged ideal MHD equations for the mixture of thermal and nonthermal solar wind ions coupled with the kinetic Boltzmann equation describing the transport of neutral atoms. An adaptive mesh refinement technique can be employed for efficient high-resolution calculations. The MS-FLUKSS heliospheric MHD model is coupled with the WSA model (Kim et al., 2020), which uses both ADAPT-GONG and ADAPT-HMI input magnetograms, with the PFSS source surface height and the WSA outer boundary at 2.5 and 10 solar radii, respectively. Hence, field line tracing is performed through the MHD domain down to 10 solar radii instantaneously at approximately 1 hour cadence, where the origin of the field line on the photosphere is already known, as described for WSA.
see Prediction and Verification of Parker Solar Probe Solar Wind Sources at 13.3Rs, Badman et al. (2023)
