Progress 07/01/23 to 02/29/24

Outputs
Target Audience:This project has the potential to help many different industries that utilize water from snow melt. Our initial target audience is the USDA, NRCS Snow Survey and Water Supply Forcasting Program. Other audiences include: Snow researchers and scientists Recreational users like skiers, snowboarders, snowmobilers, fishermen, and kayakers Forecast hydrologists in other agencies and governments (USBR, NOAA, COE, USFS, states) Agricultural Irrigators The US Army Corp of Engineers Hydroelectric power generators Municipalities who use river resources Changes/Problems:The methodology described in the Phase I proposal was sound, but the work plan was too aggressive for a 9-month Phase I. Proving outthe proposed methodrequired correctly implementingseveral very complex models and correction factors. These included picosecond-level clock corrections, multidimensional orbit state propagation, corrections for special and general relativity, crustal tide models, andantenna phase center characterizations. Implementing each component of the method took longer than anticipated and the accumulated delays prevented a full demonstation of the snow retrieval before the end ofPhase I. By the end of the project, it was apparent that bugs in both the phase wind up correction and the earth tide modelwere causing systematically repeating errors on the order of~50 mm. Without identifying the cause of these errors, both the location solver and the snow solver could not be fully tested. (This level of error prevents a meaningful snow retrieval.)Therefore, the snow retrieval algorithm was not proven out as originally intended. What opportunities for training and professional development has the project provided? Nothing Reported How have the results been disseminated to communities of interest? Nothing Reported What do you plan to do during the next reporting period to accomplish the goals? Nothing Reported

Impacts
What was accomplished under these goals? Accomplishments are grouped by goal below: Sophisticated algorithms were implemented in both Python and C++ which process the GNSS observables as follows: Software was written which downloads and applies precise orbital stateand clock correction factors to the GNSS carrier phase data. Earth tide models were implemented in both Python and C++. These correct for the tidal displacement of the Earth's crust due to the moon and the sun. Tropospheric delay models were implemented in both Python and C++. A key objective was met with regard to the zenith wet delay modeling. A global model of the zenith wet delay based on surface pressure, humidity, and temperature was built using a neural network. This model was substantially better than previously published models and predictsprecipitable water vapor to ~1 mm (1-ð) in winter conditions. This model will allow the snow retrieval algorithm to compensate for delays in the atmosphere which behave similarly to the delays induced by the snowpack. Typical GNSS corrections including those whichmodel the Sagnac effect, special relativity, general relativity, and phase wind up were implemented. A method of deducing the carrier phase bias was successfully implemented. Draft implementations of the location and snow solvers were developed in Python and C++. We were unable to fully test theseprior to the end of the project. Large portions of our previous C++ source code base were refactored to allow common code to be shared between the two-receiver and single-receiver systems. Python and bash scripts were written that automate post-processing and generation of figure plots for instrument intercomparison. Due to issues with the algorithm testing, comparison between SWE data products from the proposed method and existing technologies were not made. Due to unforeseen issues encountered during algorithm development, no accuracy characterizations were made prior to the end ofthe Phase I. Our previous processing board was adapted to include a new cellular modem and a barometric pressure sensor which is required for the wet delay model. This board ingests data from the GNSS receivers and collects snow and air temperature data. It also supports a relative humidity sensor. We sent the board for manufacturing in November 2023. After extensive testing at our home office, we replaced the existing boards at each of our three preexisting field deployment sites in Februaryto gain field experience with the new boards under winter conditions. The resulting design is intended to be a first revision of a production system. Its realization fulfills the objective of having a prototype system ready for the 2024-25 winter. We feel that having a hardware product with this level of maturity by the end of Phase I is a huge success.

Publications