Search

Climate change threatens our ability to grow food for an ever-increasing population. There is a need for high-quality soil moisture predictions in under-monitored regions like Africa. However, it is unclear if soil moisture processes are globally similar enough to allow our models trained on available in situ data to maintain accuracy in unmonitored regions. We present a multitask long short-term memory (LSTM) model that learns simultaneously from global satellite-based data and in situ soil moisture data. This model is evaluated in both random spatial holdout mode and continental holdout mode (trained on some continents, tested on a different one). The model compared favorably to current land surface models, satellite products, and a candidate machine learning model, reaching a global median correlation of 0.792 for the random spatial holdout test. It behaved surprisingly well in Africa and Australia, showing high correlation even when we excluded their sites from the training set, but it performed relatively poorly in Alaska where rapid changes are occurring. In all but one continent (Asia), the multitask model in the worst-case scenario test performed better than the soil moisture active passive (SMAP) 9 km product. Factorial analysis has shown that the LSTM model's accuracy varies with terrain aspect, resulting in lower performance for dry and south-facing slopes or wet and north-facing slopes. This knowledge helps us apply the model while understanding its limitations. This model is being integrated into an operational agricultural assistance application which currently provides information to 13 million African farmers. [ABSTRACT FROM AUTHOR]

Climate change threatens our ability to grow food for a growing population. There are concurrent droughts and floods happening globally, with the greatest impacts felt in Africa. There is a need for high-quality soil moisture predictions in under-monitored regions like Africa. Yet it is unclear if soil moisture processes are globally similar enough to allow our models to maintain accuracy in unmonitored regions. We present a multitask long short-term memory (LSTM) model that learns simultaneously from global satellite-based and in-situ soil moisture data. This model is evaluated in both random spatial holdout mode and continental holdout mode (trained on some continents, tested on a different one). The model compared favorably to current land surface models, satellite products, and a candidate machine learning model, reaching a global median correlation of 0.792 for the random spatial holdout test. It behaved surprisingly well in Africa and Australia, showing high correlation even when we excluded their sites from the training set, but performed relatively poorly in Alaska where rapid changes are occurring. In all but one continent (Asia), the multitask model in the worst-case scenario test performed better than the soil moisture active passive (SMAP) 9-km product. Factorial analysis shows that the LSTM model’s accuracy with representing impacts of terrain aspect, resulting in lower performance for dry and south-facing slopes or wet and north-facing slopes. This knowledge helps us apply the model while understanding its limitations. This model is being integrated into an operational agricultural assistance application which currently provides information to 13 million African farmers. [ABSTRACT FROM AUTHOR]

We present discrete (2‐h resolution) multi‐year (2008–2017) in situ measurements of seasonal vegetation growth and soil biophysical properties from two sites on Alaska's North Slope, USA, representing dry and wet sedge tundra. We examine measurements of vertical active soil layer temperature and soil moisture profiles (freeze/thaw status), woody shrub vegetation physiological activity, and meteorological site data to assess interrelationships within (and between) these two study sites. Vegetation phenophases (cold de‐hardening start, physiological function start, stem growth start, stem growth end, physiological function end, cold hardening completion) were found to have greater interannual day of year (DOY) occurrence variability at the dry site compared with the wet site. At the dry site, vegetation activity begins on average ~7 days earlier and ends ~11 days earlier. The mean active stem growth window lasts ~54 days for the dry site and ~51 days for the wet site. Vegetation, in both tundra environments, began cold de‐hardening functions (warm season prep) prior to atmospheric temperatures warming above 0°C. Similar results were found related to the critical soil freeze/thaw/transition dates; the dry site had a DOY phenophase occurrence range that was 8 days larger than that of the wet site. A longer continuous summer thaw period was captured at the wet site by ~26 days throughout the active layer. In addition, the dry site was measured to have longer spring and fall soil isothermal conditions than the wet site by ~9 and 5 days throughout the active layer. These results show that the dry site's willow shrub vegetation physiology and soil condition phenology is more variable than the wet site. Alongside the in situ data, a remote sensing product from NASA's MEaSUREs program was utilized; our research indicates that the AMSR‐derived satellite product is more precise over the wet tundra site with critical date alignment between remote sensing observations and in situ measurements ranging from ~4 to 11 days. Furthermore, the AMSR product was shown to preemptively estimate land surface condition change during the spring transition for both tundra types while lagging during the fall transition and freeze‐up periods. [ABSTRACT FROM AUTHOR]

Recent warming at high-latitudes has accelerated permafrost thaw in northern peatlands, and thaw can have profound effects on local hydrology and ecosystem carbon balance. To assess the impact of permafrost thaw on soil organic carbon (OC) dynamics, we measured soil hydrologic and thermal dynamics and soil OC stocks across a collapse-scar bog chronosequence in interior Alaska. We observed dramatic changes in the distribution of soil water associated with thawing of ice-rich frozen peat. The impoundment of warm water in collapse-scar bogs initiated talik formation and the lateral expansion of bogs over time. On average, Permafrost Plateaus stored 137 ± 37 kg C m, whereas OC storage in Young Bogs and Old Bogs averaged 84 ± 13 kg C m. Based on our reconstructions, the accumulation of OC in near-surface bog peat continued for nearly 1,000 years following permafrost thaw, at which point accumulation rates slowed. Rapid decomposition of thawed forest peat reduced deep OC stocks by nearly half during the first 100 years following thaw. Using a simple mass-balance model, we show that accumulation rates at the bog surface were not sufficient to balance deep OC losses, resulting in a net loss of OC from the entire peat column. An uncertainty analysis also revealed that the magnitude and timing of soil OC loss from thawed forest peat depends substantially on variation in OC input rates to bog peat and variation in decay constants for shallow and deep OC stocks. These findings suggest that permafrost thaw and the subsequent release of OC from thawed peat will likely reduce the strength of northern permafrost-affected peatlands as a carbon dioxide sink, and consequently, will likely accelerate rates of atmospheric warming. [ABSTRACT FROM AUTHOR]

Projected increases in air temperature and precipitation due to climate change in Arctic wetlands could dramatically affect ecosystem functioning. As a consequence, it is important to define the controls on evapotranspiration, which is the major pathway of water loss from these systems. We quantified the multi-year controls on midday arctic coastal wetland evapotranspiration measured with the eddy covariance method at two vegetated drained thaw lake basins near Barrow, Alaska. Variations in near-surface soil moisture and atmospheric vapor pressure deficits were found to have nonlinear effects on midday evapotranspiration rates. Vapor pressure deficits near and above 0.3 kPa appeared to be an important hydrological threshold, allowing latent heat fluxes to persistently exceed sensible heat fluxes. Dry soils increased the bulk surface resistance (water-limited). Wet soils favored ground heat flux and therefore limited the energy available to sensible and latent heat fluxes (energy-limited). Thus, midday evapotranspiration was suppressed on both dry and wet soils through different mechanisms. We also found that wet soils (ponding excluded) combined with large atmospheric demands resulted in an increased bulk surface resistance and therefore suppressing the evapotranspiration to below its potential rate (Priestley-Taylor α < 1.26). This is likely caused by the limited ability of mosses to transfer moisture during large atmospheric demands. Ultimately, in addition to net radiation, the various controlling factors on midday evapotranspiration (near-surface soil moisture, atmospheric vapor pressure, and the limited ability of mosses that are saturated at depth to transfer water during high atmospheric vapor demands) resulted in an average evapotranspiration rate of up to 75% of the potential evapotranspiration rate. The multiple limitations on midday evapotranspiration rates have the potential to moderate interannual variation of total evapotranspiration and dampen excessive water loss during a warmer climate. Combined with the prevailing maritime winds and the projected increase in precipitation, these dampening mechanisms will likely prevent extensive future soil drying and hence maintain the presence of coastal wetlands. [ABSTRACT FROM AUTHOR]

Climate warming is expected to have a large impact on plant species composition and productivity in northern latitude ecosystems. Warming can affect vegetation communities directly through temperature effects on plant growth and indirectly through alteration of soil nutrient availability. In addition, warming can cause permafrost to thaw and thermokarst (ground subsidence) to develop, which can alter the structure of the ecosystem by altering hydrological patterns within a site. These multiple direct and indirect effects of permafrost thawing are difficult to simulate in experimental approaches that often manipulate only one or two factors. Here, we used a natural gradient approach with three sites to represent stages in the process of permafrost thawing and thermokarst. We found that vascular plant biomass shifted from graminoid-dominated tundra in the least disturbed site to shrub-dominated tundra at the oldest, most subsided site, whereas the intermediate site was co-dominated by both plant functional groups. Vascular plant productivity patterns followed the changes in biomass, whereas nonvascular moss productivity was especially important in the oldest, most subsided site. The coefficient of variation for soil moisture was higher in the oldest, most subsided site suggesting that in addition to more wet microsites, there were other microsites that were drier. Across all sites, graminoids preferred the cold, dry microsites whereas the moss and shrubs were associated with the warm, moist microsites. Total nitrogen contained in green plant biomass differed across sites, suggesting that there were increases in soil nitrogen availability where permafrost had thawed. [ABSTRACT FROM AUTHOR]