Random Forest Based Downscaling of SMAP L4 Soil Moisture and Subsequent Use in Landslide Slope Stability Analysis

Daniel M. Francis, L. Sebastian Bryson

Research output: Contribution to journalConference articlepeer-review

Abstract

Soil moisture has a strong influence on the stability of shallow colluvial hillslopes. Thus, it is often necessary to monitor seasonal variations of soil moisture at individual sites of interest. This work demonstrated a workflow showing that satellite-based soil moisture data can be downscaled through machine learning and used as a source of data for landslide stability analyses. In particular, the NASA Soil Moisture Active Passive (SMAP) Level 4 Root Zone Soil Moisture (L4_SM) product provides near-real-time measurements of soil moisture on a gridded spatial resolution of 9 km. Such resolution, however, is too coarse for soil moisture measurements at a single site. Therefore, satellite-based soil moisture data must be downscaled for use in slope stability assessment at individual sites. This study used a machine learning (ML) technique, Random Forest Regression, to downscale the SMAP Level 4 Root Zone Soil Moisture product and applied the results to landslide slope stability analyses. We modeled the L4_SM data based on 1 km NASA Moderate Resolution Imaging Spectroradiometer (MODIS) products. First, the MODIS products were upscaled from 1 km to 9 km and used as features, combining with L4_SM data as labels to train the Random Forest model. Then, the L4_SM data were downscaled from 9 km to 1 km through the trained model and locally calibrated to that of ground-based data by linear regression. The processed SMAP data were served as the input of an infinite-slope stability model, which was used to investigate incipient slope failure conditions. The slope stability results were verified using known landslide occurrences in the Commonwealth of Kentucky. Results of this work showed that the stability models constructed using downscaled L4_SM data functioned well at detecting incipient conditions at the investigated sites. The purpose behind this study was to develop a process flow routine through which SMAP L4_SM data can be downscaled through machine learning as well as investigating the potential efficacy of utilizing downscaled L4_SM to conduct landslide stability analyses.

Original languageEnglish
Pages (from-to)1-11
Number of pages11
JournalGeotechnical Special Publication
Volume2022-March
Issue numberGSP 336
DOIs
StatePublished - 2022
Event2022 GeoCongress: State of the Art and Practice in Geotechnical Engineering - Advances in Monitoring and Sensing; Embankment, Slopes, and Dams; Pavements; and Geo-Education - Charlotte, United States
Duration: Mar 20 2022Mar 23 2022

Bibliographical note

Funding Information:
This material is based upon work primarily supported by the National Science Foundation (NSF) under award number EEC-1449501. Any opinions, findings and conclusions, or recommendations expressed in this paper are those of the author(s) and do not necessarily reflect those of NSF.

Funding Information:
This research was partially funded by National Aeronautics and Space Administration (NASA) under grant No. 18 -DISASTER18 -0022. The original field deployment was conducted under the Auspices of the ASCE Geo-Institute Technical Committee on Embankments, Dams and Slopes.

Funding Information:
The authors wish to express their gratitude for the funding and support provided by the Texas Department of Transportation that made this research possible.

Funding Information:
This development project was awarded as an MnDOT contract (no. 34287) [Federal Project Number TPF -5 (341)] and funded by National Road Research Alliance (NRRA). We acknowledge all those groups and organizations contributed to the fund. We thank all Technical Advisory Panel (TAP) members for the constructive feedback. Special thanks go to Jason Richter (JR) and Rebecca Embacher at MnDOT for providing insights to the proper steering of the development and also serving as liaison (JR) during the project execution.

Funding Information:
This NSF-GOALI project is funded by the National Science Foundation (NSF) under grant CMMI 1917168. The authors gratefully acknowledge our colleagues at the Los Angeles Metro (Metro) for their partnership in this research study, our colleagues who work in consulting capacities for the LA Metro and its expansion projects, as well as the LA Metro Tunnel Advisory Panel (TAP) for their continued technical input and support. Specifically, we acknowledge Dr. Androush Danielians and Mr. Joe Demello, as well as Mrs . Amanda Elioff and Dr. Roy Cook. The original design, monitoring, and field data analysis during construction of the Exploratory Shaft was conducted in part by Parsons Brinkerhoff (now WSP USA) and Wood Inc (Formerly AMEC). The Exploratory Shaft was const ructed by ICS, Inc.

Funding Information:
This study is supported by Christian R. and Mary F. Lindback Foundation and the Geotechnical Women Faculty Program (GTWF). The team would like to thank undergraduate students, Mr. Adam King, Mr. Piero Benites, and Mr. Dino Spinelli, for supporting the development of hands-on activities.

Funding Information:
The authors would like to thank the Office of Naval Research for funding this project through grant N00014-18-1-2435. The authors also thank Julia DiLeo and Raymond Turner of CSTARS for assistance with image collection as well as the technical staff at the USACE-FRF including Heidi Wadman, Jesse McNinch, and Pat Dickhut for assistance with site access and data collection. Finally, the authors would like to thank Nick Brilli assistance in data collection.

Funding Information:
Formation of a Dispersed Soil Arch in Embankm ents Supported by Columns with Cap Beams and the Development of System Efficacy ...................................221

Funding Information:
This material is based upon work supported by the National Science Foundation (NSF) under NSF Award Number CMMI–1632963. Any opinions, findings and conclusion, or recommendations expressed in this material are those of the authors, and do not necessarily reflect those of the NSF.

Funding Information:
The authors would like to acknowledge the financial assistance from the Australian Research Council (ARC -DP180101916). The assistance provided by industry (RMS, Douglas Partners ASMS, South 32, and Tyre Crumbs Australia) in relation to the procurement of m aterial used in this study is gratefully acknowledged. Some figures in this paper have been modified and reproduced from ASCE J. Geotech. Geoenviron. Engin. and J. Mater. Civil Engin.

Funding Information:
The authors would like to thank Texas Department of Transportation (TxDOT) for funding this research under project 0-6936 through the Center f or Transportation Research (CTR) at the University of Texas at Austin.

Funding Information:
This material is based upon work supported by the Delaware Department of Transportation under DelDOT Project Number T201966002, Task 1891 -28. The authors would like to express their gratitude to the Delaware Department of Transportation and Greggo & Ferrara, Inc (especially Nicholas Ferrara III and Dave Patrick) for facilitating access to the project site during the duration of this study.

Funding Information:
The National Science Foundation funded the Geotechnical Women Faculty (GTWF) Project in 2016 to promote gender parity amongst geotechnical engineering professors in the United States. The GTWF Project has, as part of its efforts, built a database of tenured and tenure -track faculty. This database was first created in 2016 at the start of the GTWF project, and recently XSGDWHG E\ WKH WHDP LQ 7KH UHVXOWV DUH ³VQDSVKRWV´ RI WKH ILHOG DW WZR SRLQWV LeQ2 W0LP6 DQG WKDW FDQ WKHQ EH FRPSDUHG WR DVVHVV WKH HIIHFWV RI LQLWLDWLYHV WR LQFUHDVH WKH ILHOG¶V gender diversity. The data also allow the status of US -based female faculty in the geotechnical field specifically to be compared to broader trends in Civil Engineering as a whole. Further, data are presented regarding inclusion in the field of geotechnical engineering through a look at IHPDOH IDFXOW\¶V VKDUH RI $6&( DZDUGV FRPPLWWHH FKDLUV DQG XQLYHUVLW\ GLVWLQJXLVKHG OHFWXUH series. Initial results show that the field of geotechnical engineering still lags behind when compared to civil engineering as a whole, and that placement of women in tenure -track faculty positions has grown only slightly since 2016. Participation data paints a picture of a field starting to see the recognition of women in the field, with female faculty receiving a higher percentage of awards for those under 35, and increasing numbers of female tenure -tracked faculty. These data indicate that efforts over the past decade to increase gender parity are meeting with real -world impact that has seen the field add female faculty at a rate well above the average for Civil (QJLQHHULQJ DV D ZKROH EXW WKDW PRUH QRZ QHHGV WR EH GRQH WR VXSSRUW ZRPHQ¶V UHFRJQLWLRQ DQG inclusion in the field. A follow -up survey in 2025 is also proposed to further assess the changing state of the field.

Funding Information:
While the widespread nature of these discussions over the past year are welcome, the discussions themselves, about the overrepresentation of white men in STEM and attempts to increase diversity, are nothing new (for example: Bhatia, 1989) . In 1989, the National Science Foundation (NSF) sponsored a workshop of all fema le JHRWHFKQLFDO IDFXOW\ IRU GLVFXVVLRQV RQ VXSSRUWLQJ DQG LQFUHDVLQJ ZRPHQ¶V UHSUHVHQWDWLRQ LQ WKH field. There were seven women geotechnical faculty present. Through the 90s and 2000s, in line with general trends in the academy and society more broadly, t he number of women in the geotechnical field steadily increased. While the number of women in the field has certainly increased in the decades since, there is no central count of geotechnical faculty demographics. The American Society for Engineering Education (ASEE) publishes yearly snapshots assessing the demographics of engineering subfields, but geotechnical, as a specialty of the subfield civil engineering, is not among them. Three hundred and eight universities in the United States offer Civil Enginee ring degrees ( Dept. of Education, n.d.) , out of which we identified 181 that employ tenure or tenure -tracked faculty in the subfield of geotechnical engineering. U.S. institutions graduated 17,102 students in the year 2016 in the field of civil engineering, increasing to 20,056 just four years later in 2020. Of those 2016 graduates, 24.4% were female; LQ ZRPHQ¶V VKDUH KDG LQFUHDVHG WR MXVW RYHU D TXDUWHU DW 6LPLODU JURZWK ZDV SUHVHQW LQ ZRPHQ¶V UHSUHVHQWDWLRQ LQ FLYLO HQJLQHHULQJ DPRQJVW WHQXUHG RU WHQX-UtHrack faculty: from 18.4% in 2016 to 19.6% in 2020 ³(QJLQHHULQJ E\ WKH 1XPEHUV ´ ? <RGHUA s we have demonstrated previously (S. K. Bhatia et al., 2021) , increasing gender diversity among faculty in the field leads to increased diversity among students, and practicing engineers, by GLVUXSWLQJ WKH ³IO\ZKHHO´ HIIHFW RI JHQGHU GLVFULPLQDWL(sRQe e also: Laefer et al., 2007) . In 2016, the Geotechnical Women Faculty (GTWF) Project received funding from the NSF an aim to lowe" r isolation of female faculty in geotechnical engineering by creating and sustaining a supportive national network that driv HV FDUHHU VXFFHVV LQ DFDGHPLD´(GTWF, 2016) . The GTWF project pursued this by offering seed grants to women faculty, which required first the identification of eligible faculty ²the creation of the first U.S. Geotechnical faculty database in 2016 (Alestalo et al., 2015; P. Gallagher et al., 2019 ; P. M. Gallagher et al., 2018; D. F. Laefer & McHale, 2010) . In 2018 the database was refreshed, with new faculty identified and faculty ranks adjusted. The same process was repeated in 2021, giving us the ability to spot changes and growth in geotechnic al demographics, numbers, and rank. This paper will discuss the 2021 database, and with the changes observed between 2016 and 2021. We will go over the methodology that was employed to create a directory database of U.S. faculty, and introduce some shortcomings and caveats around the data before we discuss the UHVXOWLQJ ³VQDSVKRWV´ LQ GHSWK ,Q DGGLWLRQ WR SUHVHQWLQJ D ³6WDWH RI WKH )LHOG´ ZLWK UHJDUGV WR JHQGHU GLYHUVLW\ ZH KDYH collected historical data for awards, prestigious university lecture series, and editorial positions for the field from 1990 through 2021. This data we are using as a proxy measurement for ZRPHQ¶V UHFRJQLWLRQ LQ WKH ILHOG RI JHRWHFKQLFDO HQJLQHHULQJ 6R WKDW ZKHQ FKHFNHG DJDLQVW demographic data, the percentage of women who are re cognized by awards or in positions of KLJK YLVLELOLW\ ZRXOG DSSUR[LPDWH ZRPHQ¶V RYHUDOO GHPRJUDSKLF SUHVHQFH DQG UHFRJQLWLRQ

Funding Information:
Support for this study provided by Bay Area Rapid Transit (BART) under agreement 6M3466 is gratefully acknowledged. The authors thank Chung-Soo Doo and Carlos Rosales of BART for their support. The authors also thank Andrew Sander, Noah Aldrich, Michael Sanders and Darren McKay for their input and support during construction and testing of the specimen.

Funding Information:
A series of graduate students have contributed to this research program on pipe joints, with the experimental work of Anderson de Oliveira, Eric Poon, and Min Zhou used in the current paper. The contributions to this research program made by David Becerril García, Graeme Boyd, Josh Coughlan, Richard Foley, Tom Fromberger, Haitao Lan, Xiaogang Qin, Masoumeh Saiyar, Andy Take, Yu Wang and Rui Yong are also greatly appreciated. The tests presented here were funded using Discovery and Strategic Research Grants provided by the Natural Sciences and Engineering Research Council of Canada. Some pipe specimens were donated by Armtec Industries.

Funding Information:
This research was sponsored by the National Science Foundation (Award #1928813) with counterpart funding from the Iowa Highway Research Board.

Funding Information:
This work is supported by the Geosynthetic Institute with the 2020 grant, Folsom, PA [32 -4116 -65048, Louisiana Tech University] and by Transportation Center, Norman, Ok [SPTC15.1 -23].

Funding Information:
This research was conducted as an additional follow-up study to efforts partially supported by the California Department of Transportation (Caltrans) under Contract No. 65A0548 with Dr. Charles Sikorsky as the project manager.

Funding Information:
The authors would like to express their gratitude for the funding and support provided by Texas Department of Transportation that made this research possible.

Publisher Copyright:
© ASCE

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Architecture
  • Building and Construction
  • Geotechnical Engineering and Engineering Geology

Fingerprint

Dive into the research topics of 'Random Forest Based Downscaling of SMAP L4 Soil Moisture and Subsequent Use in Landslide Slope Stability Analysis'. Together they form a unique fingerprint.

Cite this