Upper limits for post-wildfire floods and distinction from debris flows.

Upper magnitude limits and scaling with basin size for post-wildfire floods are unknown. An envelope curve was estimated defining post-wildfire flood upper limits as a function of basin area. We show the importance of separating peak flows by floods versus debris flows. Post-wildfire flood maxima are a constant 43 m3 s-1 km-2 for basins from 0.01 to 23 to 34 km2 and then declining with added basin area according to a power law relation. Intense rainfall spatial scaling may cause the envelope curve threshold at 23 to 34 km2. Post-wildfire flood maxima are smaller than unburned flood maxima for similar basin area. Rainstorm comparisons indicate that post-wildfire floods are triggered by smaller precipitation depths than unburned floods. Post-wildfire exceptional floods are driven by extreme rainfall rates, in contrast to post-wildfire debris flows. Runoff rates for post-wildfire envelope floods are consistent with infiltration-excess runoff. Future increases in precipitation intensity or wildfire frequency and extent could increase post-wildfire flood upper limits.

Post-fire temporal trends in soil-physical and -hydraulic properties and simulated runoff generation: Insights from different burn severities in the 2013 Black Forest Fire, CO, USA.

Burn severity influences on post-fire recovery of soil-hydraulic properties controlling runoff generation are poorly understood despite the importance for parameterizing infiltration models. We measured soil-hydraulic properties of field-saturated hydraulic conductivity (Kfs), sorptivity (S), and wetting front potential (ψf) for four years after the 2013 Black Forest Fire, Colorado, USA, at six sites across a gradient of initial remotely sensed burn severity using the change in the normalized burn ratio (dNBR). These measurements were correlated with soil-physical property measurements of bulk density (ρb), loss on ignition (LOI, a measure of soil organic matter), and ground cover composition to provide insight into causal factors for temporal changes in Kfs, S, and ψf. Modeled infiltration using the Smith-Parlange approach parameterized with measured Kfs, S, and ψf further discerned the role of precipitation intensity on runoff generation. Temporal trends of soil-physical properties and ground cover showed influences from initial burn severity. Trends in soil-hydraulic properties, surprisingly, were not strongly influenced by initial burn severity despite inferred effects of ρb, LOI, and ground cover on trends in Kfs and S. Calculations of dNBR at the time of sampling showed strong correlations with Kfs and S, demonstrating a new approach for estimating long-unburned Kfs and S values, infiltration model parameters after fire, and assessing the time of return to pre-fire values. Simulated infiltration-excess runoff, in contrast, did depend on initial burn severity. Time series of the ratio S2/Kfs ≈ ψf tended to converge between 1 and 10 mm four years after wildfire, potentially (i) defining a long-unburned forest domain of S2/Kfs and ψf from 1 to 10 mm with relatively high Kfs values, and (ii) providing a new post-fire soil-hydraulic property recovery metric (i.e. S2/Kfs ≈ ψf in the range of 1 to 10 mm) for sites in the Rocky Mountains of the USA.

Fates and fingerprints of sulfur and carbon following wildfire in economically important croplands of California, U.S.

Sulfur (S) is widely used in agriculture, yet little is known about its fates within upland watersheds, particularly in combination with disturbances like wildfire. Our study examined the effects of land use and wildfire on the biogeochemical "fingerprints," or the quantity and chemical composition, of S and carbon (C). We conducted our research within the Napa River Watershed, California, U.S., where high S applications to vineyards are common, and ~ 20% of the watershed burned in October 2017, introducing a disturbance now common across the warmer, drier Western U.S. We used a laboratory rainfall experiment to compare unburned and low severity burned vineyard and grassland soils. We then sampled streams draining sub-catchments with differing land use and degrees of burn and burn severity to understand combined effects at broader spatial scales. Before the laboratory experiment, vineyard soils had 2-3.5 times more S than grassland soils, while burned soils-regardless of land use-had 1.5-2 times more C than unburned soils. During the laboratory experiment, vineyard soil leachates had 16-20 times more S than grassland leachates, whereas leachate C was more variable across land use and burn soil types. Unburned and burned vineyard soils leached S with δ34S values enriched 6-15‰ relative to grassland soils, likely due to microbial S processes within vineyard soils. Streams draining vineyards also had the fingerprint of agricultural S, with ~2-5 fold higher S concentrations and ~ 10‰ enriched δ34S-SO42- values relative to streams draining non-agricultural areas. However, streams draining a higher fraction of burned non-agricultural areas also had enriched δ34S values relative to unburned non-agricultural areas, which we attribute to loss of 32S during combustion. Our findings illustrate the interacting effects of wildfire and land use on watershed S and C cycling-a new consideration under a changing climate, with significant implications for ecosystem function and human health.