Laurel G. Larsen

Laurel G. Larsen

Auto-generated profile, based on publicly available information

Laurel G. Larsen

University of California Berkeley Berkeley CA USA

Research Focus & Metrics

Academic Career Length

18Years (2009-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
25 December 2020 in Water
Verified
Journal Article
Open access
PDF file
EMEdom Moges
Edom Moges
YDYonas Demissie
Yonas Demissie
LLLaurel Larsen
Laurel Larsen
Fuad YassinFuad Yassin
Fuad Yassin
Despite progresses in representing different processes, hydrological models remain uncertain. Their uncertainty stems from input and calibration data, model structure, and parameters. In characterizing these sources, their causes, interactions and different uncertainty analysis (UA) methods are reviewed. The commonly used UA methods are categorized into six broad classes: (i) Monte Carlo analysis, (ii) Bayesian statistics, (iii) multi-objective analysis, (iv) least-squares-based inverse modeling, (v) response-surface-based techniques, and (vi) multi-modeling analysis. For each source of uncertainty, the status-quo and applications of these methods are critiqued in gauged catchments where UA is common and in ungauged catchments where both UA and its review are lacking. Compared to parameter uncertainty, UA application for structural uncertainty is limited while input and calibration data uncertainties are mostly unaccounted. Further research is needed to improve the computational efficiency of UA, disentangle and propagate the different sources of uncertainty, improve UA applications to environmental changes and coupled human–natural-hydrologic systems, and ease UA’s applications for practitioners.

26 April 2018 in Geophysical Research Letters
Verified
Journal Article
Open access
PDF file
MLM. C. Levy
M. C. Levy
ALA. V. Lopes
A. V. Lopes
ACAvery Cohn
Avery Cohn
LLL. G. Larsen
L. G. Larsen
STS. E. Thompson
S. E. Thompson
Nearly half of recent decades' global forest loss occurred in the Amazon and Cerrado (tropical savanna) biomes of Brazil, known as the arc of deforestation. Despite prior analysis in individual river basins, a generalizable empirical understanding of the effect of deforestation on streamflow across this region is lacking. We frame land use change in Brazil as a natural experiment and draw on in situ and remote sensing evidence in 324 river basins covering more than 3 × 106 km2 to estimate streamflow changes caused by deforestation and agricultural development between 1950 and 2013. Deforestation increased dry season low flow by between 4 and 10 percentage points (relative to the forested condition), corresponding to a regional‐ and time‐averaged rate of increase in specific streamflow of 1.29 mm/year2, equivalent to a 4.08 km3/year2 increase, assuming a stationary climate. In conjunction with rainfall and temperature variations, the net (observed) average increase in streamflow over the same period was 0.76 mm/year2, or 2.41 km3/year2. Thus, net increases in regional streamflow in the past half century are 58% of those that would have been experienced with deforestation given a stationary climate. This study uses a causal empirical analysis approach novel to the water sciences to verify the regional applicability of prior basin‐scale studies, provides a proof of concept for the use of observational causal identification methods in the water sciences, and demonstrates that deforestation masks the streamflow‐reducing effects of climate change in this region.

15 July 2017 in Water Resources Research
Verified
Journal Article
Open access
PDF file
CTChristopher J. Tennant
Christopher J. Tennant
AHAdrian A. Harpold
Adrian A. Harpold
KLKathleen Ann Lohse
Kathleen Ann Lohse
SGSarah E. Godsey
Sarah E. Godsey
BCBenjamin T. Crosby
Benjamin T. Crosby
NGNancy Glenn
Nancy Glenn
In mountains with seasonal snow cover, the effects of climate change on snowpack will be constrained by landscape‐vegetation interactions with the atmosphere. Airborne lidar surveys used to estimate snow depth, topography, and vegetation were coupled with reanalysis climate products to quantify these interactions and to highlight potential snowpack sensitivities to climate and vegetation change across the western U.S. at Rocky Mountain (RM), Northern Basin and Range (NBR), and Sierra Nevada (SNV) sites. In forest and shrub areas, elevation captured the greatest amount of variability in snow depth (16–79%) but aspect explained more variability (11–40%) in alpine areas. Aspect was most important at RM sites where incoming shortwave to incoming net radiation (SW:NetR↓) was highest (∼0.5), capturing 17–37% of snow depth variability in forests and 32–37% in shrub areas. Forest vegetation height exhibited negative relationships with snow depth and explained 3–6% of its variability at sites with greater longwave inputs (NBR and SNV). Variability in the importance of physiography suggests differential sensitivities of snowpack to climate and vegetation change. The high SW:NetR↓ and importance of aspect suggests RM sites may be more responsive to decreases in SW:NetR↓ driven by warming or increases in humidity or cloud cover. Reduced canopy‐cover could increase snow depths at SNV sites, and NBR and SNV sites are currently more sensitive to shifts from snow to rain. The consistent importance of aspect and elevation indicates that changes in SW:NetR↓ and the elevation of the rain/snow transition zone could have widespread and varied effects on western U.S. snowpacks.

1 June 2019 in Geomorphology
Verified
Journal Article
Closed access
LLLaurel G. Larsen
Laurel G. Larsen

19 August 2020 in Water Resources Research
Verified
Journal Article
Open access
PDF file
CTChristopher Tennant
Christopher Tennant
LLLaurel Larsen
Laurel Larsen
DBDino Bellugi
Dino Bellugi
EMEdom Moges
Edom Moges
LZLiang Zhang
Liang Zhang
HMHongxu Ma
Hongxu Ma
Streamflow forecasts often perform poorly because of improper representation of hydrologic response timescales in underlying models. Here, we use transfer entropy (TE), which measures information flow between variables, to identify dominant drivers of discharge and their timescales using sensor data from the Dry Creek Experimental Watershed, ID, USA. Consistent with previous mechanistic studies, TE revealed that snowpack accumulation and partitioning into melt, recharge, and evaporative loss dominated discharge patterns and that snow‐sourced baseflow reduced the greatest amount of uncertainty in discharge. We hypothesized that machine learning models (MLMs) specified in accordance with the dominant lag timescales, identified via TE, would outperform timescale‐agnostic models. However, while lagged‐variable random forest regressions captured the dominant process—seasonal snowmelt—they ultimately did not perform as well as the unlagged models, provided those models were specified with input data aggregated over a range of timescales. Unlagged models, not constrained by timescales of the dominant processes, more effectively represented variable interactions (e.g., rain‐on‐snow events) playing a critical role in translating precipitation into streamflow over long, intermediate, and short timescales. Meanwhile, long short‐term memory (LSTM) models were effective in internally identifying the key lag and aggregation scales for predicting discharge. Parsimonious specification of LSTM models, using only daily unlagged precipitation and temperature data, produced the highest performing predictions. Our findings suggest that TE can identify dominant streamflow controls and the relative importance of different mechanisms of streamflow generation, useful for establishing process baselines and fingerprinting watersheds. However, restricting MLMs based on dominant timescales undercuts their skill at learning these timescales internally.

12 August 2014 in Eos
Verified
Journal Article
Open access
PDF file
LLLaurel Larsen
Laurel Larsen
CTChris Thomas
Chris Thomas
MEMaarten Eppinga
Maarten Eppinga
TCTom Coulthard
Tom Coulthard
On 22 May 2011 a massive tornado tore through Joplin, Mo., killing 158 people. With winds blowing faster than 200 miles per hour, the tornado was the most deadly in the United States since modern record keeping began in the 1950s.