diff --git a/analysis/plot_radiation_sites_biomes.R b/analysis/plot_radiation_sites_biomes.R new file mode 100644 index 0000000..ddf25ef --- /dev/null +++ b/analysis/plot_radiation_sites_biomes.R @@ -0,0 +1,307 @@ +library(dplyr) +library(tidyr) +library(purrr) +library(readr) +library(lubridate) +library(hms) +library(khroma) +library(ggplot2) +library(cowplot) + +source(here::here("R/read_fdk.R")) + +# Reading ----------- +# read from file +# df_sites <- read_rds(here::here("data/df_sites.rds")) +df_sites <- read_csv(here::here("data/fdk_site_info.csv")) + +# chose representative sites +use_sites <- c( + "FI-Hyy", # Boreal Forests/Taiga + "DE-Hai", # Temperate Broadleaf & Mixed Forests + "DE-Tha", # Temperate Coniferous + "DE-Gri", # Grassland just next to DE-Tha + "BR-Sa3", # Tropical + "US-ICh" # Tundra +) + +# subset sites +df_sites_sub <- df_sites |> + filter(sitename %in% use_sites) + +# saveRDS(df_sites_sub, file = here::here("data/df_sites_sub.rds")) + +## Half-hourly data -------------- +# read hourly flux data for each site +hdf <- df_sites_sub |> + dplyr::select(sitename, lon, lat, elv) |> + mutate(data = purrr::map( + sitename, ~read_fdk( + ., + path = "~/data/FluxDataKit/v3.1/fluxnet/", + # path = "~/data/FluxDataKit/FLUXDATAKIT_FLUXNET", + pattern = "HH" + ))) |> + unnest(data) |> + group_by(sitename) |> + nest() + +## Daily data -------------- +# read daily flux data for each site +ddf <- df_sites_sub |> + select(sitename, lon, lat, elv) |> + mutate(data = purrr::map( + sitename, ~read_fdk( + ., + path = "~/data/FluxDataKit/v3.1/fluxnet/", + # path = "~/data/FluxDataKit/FLUXDATAKIT_FLUXNET", + pattern = "DD" + ))) |> + unnest(data) |> + group_by(sitename) |> + nest() + +# Aggregate ----------- +## Hourly ----------- +# aggregate to hours for mid-summer +hdf_sub <- hdf |> + unnest(data) |> + mutate(TIMESTAMP_START = ymd_hm(TIMESTAMP_START), + TIMESTAMP_END = ymd_hm(TIMESTAMP_END)) |> + mutate(year = lubridate::year(TIMESTAMP_START), + month = lubridate::month(TIMESTAMP_START), + day = lubridate::mday(TIMESTAMP_START)) |> + filter(month == 7) |> + mutate(hm = as_hms(TIMESTAMP_START)) |> + group_by(sitename, hm) |> + summarise( + # latent heat flux + le = mean(LE_F_MDS, na.rm = TRUE), + + # sensible heat flux + heat = mean(H_F_MDS, na.rm = TRUE), + + # net radiation + netrad = mean(NETRAD, na.rm = TRUE), + + # shortwave radiation + sw = mean(SW_IN_F_MDS, na.rm = TRUE), + + # longwave radiation + lw = mean(LW_IN_F_MDS, na.rm = TRUE) + ) + +## Daily to mean season ----------- +# aggregate to mean seasonal cycle +ddf_sub <- ddf |> + unnest(data) |> + mutate(TIMESTAMP = ymd(TIMESTAMP)) |> + mutate(year = year(TIMESTAMP), + doy = yday(TIMESTAMP)) |> + group_by(sitename, doy) |> + summarise( + # latent heat flux + le = mean(LE_F_MDS, na.rm = TRUE), + + # sensible heat flux + heat = mean(H_F_MDS, na.rm = TRUE), + + # net radiation + netrad = mean(NETRAD, na.rm = TRUE), + + # shortwave radiation + sw = mean(SW_IN_F_MDS, na.rm = TRUE), + + # longwave radiation + lw = mean(LW_IN_F_MDS, na.rm = TRUE) + ) |> + mutate(date = ymd("2023-01-01") + doy - 1) + + +# Plot ----------- +## Diurnal cycle in July ----------- +### Components ---------------- +hdf_sub |> + mutate(res = heat + le - netrad) |> + pivot_longer(cols = c(netrad, heat, le)) |> # , res + mutate(sitename = factor( + sitename, + levels = c( + "US-ICh", # Tundra + "FI-Hyy", # Boreal Forests/Taiga + "BR-Sa3", # Tropical + "DE-Hai", # Temperate Broadleaf & Mixed Forests + "DE-Tha", # Temperate Coniferous + "DE-Gri" # Grassland just next to DE-Tha + ) + ) + # name = factor(name, levels = c("netrad", "heat", "le", "res")) + ) |> + ggplot(aes(hm, value, color = name)) + + geom_hline(yintercept = 0, linetype = "dotted") + + geom_line() + + scale_x_time() + + # scale_x_datetime(date_breaks= "2 hours", date_labels = "%H:%M") + + scale_color_manual( + name = "", + labels = c( + heat = expression(paste(italic("H"))), + le = expression(paste(lambda, italic("E"))), + netrad = expression(paste(italic("R")[net])) + # res = "Residual" + ), + values = c( + heat = "#E69F00", + le = "#56B4E9", + netrad = "black" + # res = "grey70" + ) + ) + + # scale_linetype_manual( + # name = "", + # labels = c( + # heat = expression(paste(italic("H"))), + # le = expression(paste(lambda, italic("E"))), + # netrad = expression(paste(italic("R")[net])), + # res = "Residual" + # ), + # values = c( + # heat = "solid", + # le = "solid", + # netrad = "solid", + # res = "dashed" + # ) + # ) + + theme_classic() + + labs( + x = "Time of day", + y = expression(paste("Energy flux (W m"^-2, ")")) + ) + + facet_wrap(~sitename) + +ggsave( + here::here("book/images/diurnal_cycle_radiation.png"), + width = 12, + height = 7 +) + +### Residual (ground heat flux) ----------- +hdf_sub |> + mutate(res = heat + le - netrad) |> + mutate(sitename = factor( + sitename, + levels = c( + "US-ICh", # Tundra + "FI-Hyy", # Boreal Forests/Taiga + "BR-Sa3", # Tropical + "DE-Hai", # Temperate Broadleaf & Mixed Forests + "DE-Tha", # Temperate Coniferous + "DE-Gri" # Grassland just next to DE-Tha + ) + )) |> + ggplot(aes(hm, res)) + + geom_hline(yintercept = 0, linetype = "dotted") + + geom_line() + + scale_x_time() + + theme_classic() + + labs( + x = "Time of day", + y = expression(paste("Energy flux (W m"^-2, ")")) + ) + + facet_wrap(~sitename) + +## Mean seasonal cycle ----------- +### Components ---------- +ddf_sub |> + # mutate(res = heat + le - netrad) |> + pivot_longer(cols = c(netrad, heat, le)) |> # , res + mutate(sitename = factor( + sitename, + levels = c( + "US-ICh", # Tundra + "FI-Hyy", # Boreal Forests/Taiga + "BR-Sa3", # Tropical + "DE-Hai", # Temperate Broadleaf & Mixed Forests + "DE-Tha", # Temperate Coniferous + "DE-Gri" # Grassland just next to DE-Tha + ) + )) |> + ggplot(aes(date, value, color = name)) + + geom_hline(yintercept = 0, linetype = "dotted") + + geom_line() + + scale_x_date(date_breaks = "1 month", date_labels = "%b") + + scale_color_manual( + name = "", + labels = c( + heat = expression(paste(italic("H"))), + le = expression(paste(lambda, italic("E"))), + netrad = expression(paste(italic("R")[net])) + # res = "Residual" + ), + values = c( + heat = "#E69F00", + le = "#56B4E9", + netrad = "black" + # res = "grey70" + ) + ) + + theme_classic() + + labs( + x = "", + y = expression(paste("Energy flux (W m"^-2, ")")) + ) + + facet_wrap(~sitename) + +ggsave( + here::here("book/images/seasonal_cycle_radiation.png"), + width = 12, + height = 7 +) + +# # annual values: GPP vs PPFD +# adf_mean <- ddf |> +# unnest(data) |> +# mutate(year = year(TIMESTAMP)) |> +# group_by(sitename, year) |> +# summarise( +# gpp = sum(GPP_NT_VUT_REF), +# sw = sum(SW_IN_F_MDS) +# ) |> +# ungroup() |> +# group_by(sitename) |> +# summarise( +# gpp = mean(gpp), +# sw = mean(sw) +# ) +# +# adf_mean |> +# ggplot(aes(sw, gpp)) + +# geom_point() + + +### Residual (ground heat flux) -------- +# ground heat flux +ddf_sub |> + mutate(res = heat + le - netrad) |> + mutate(sitename = factor( + sitename, + levels = c( + "US-ICh", # Tundra + "FI-Hyy", # Boreal Forests/Taiga + "BR-Sa3", # Tropical + "DE-Hai", # Temperate Broadleaf & Mixed Forests + "DE-Tha", # Temperate Coniferous + "DE-Gri" # Grassland just next to DE-Tha + ) + )) |> + ggplot(aes(date, res)) + + geom_hline(yintercept = 0, linetype = "dotted") + + geom_line() + + scale_x_date(date_breaks = "1 month", date_labels = "%b") + + theme_classic() + + labs( + x = "", + y = expression(paste("Energy flux (W m"^-2, ")")) + ) + + facet_wrap(~sitename) + diff --git a/book/ecohydrology.qmd b/book/ecohydrology.qmd index bcd960a..7a28848 100644 --- a/book/ecohydrology.qmd +++ b/book/ecohydrology.qmd @@ -2,3 +2,9 @@ Coming soon. +- Forests sustain the hydrologic cycle through evapotranspiration, which cools climate through feedbacks with clouds and precipitation +- Tropical forests: + - "Observations show that forest transpiration is sustained during the dry season (11); this is seen also in CO2 fluxes (12) and satellite monitoring of vegetation (13, 14), to a greater extent than represented in many models." (Bonan 2008) +- Material from trial lecture: lecture_et_giub.key +- Geographic patterns of evapotranspira- tion are explained by Budyko’s analysis of the control of evapotranspiration by net radiation and precipitation. +- 3.4 Hydrologic Cycle (Bonan) \ No newline at end of file diff --git a/book/globalcarbonpatterns.qmd b/book/globalcarbonpatterns.qmd index 0336081..2bf6428 100644 --- a/book/globalcarbonpatterns.qmd +++ b/book/globalcarbonpatterns.qmd @@ -325,6 +325,14 @@ df_holo |> scale_y_continuous(expand = c(0,0)) ``` +::: {.callout-caution} +## Exercise + +1. High solar radiation during summer drives monsoonal patterns in the tropics and sub-tropics. Considering @fig-stoa_hovmoeller_holo, how do you expect monsoonal precipitation intensity in the northern part of the African continent to have changed around 6 kyr BP, compared to today. + +::: + + ## Phenology {#sec-phenology} In general, phenology refers to the study of periodic events of biological activity. In most cases, such periodic events are linked the seasonal variations of the climate. In the context of terrestrial photosynthesis and land-climate interactions, the most important phenological event is the leaf unfolding and leaf senescence and shedding in response to seasonal variations in temperature, radiation, and water availability. The presence (or absence) of active green leaves has major implications for CO~2~, water vapour, and energy exchange between the land surface and the atmosphere. @@ -333,6 +341,14 @@ The presence of green leaves is recorded by optical satellite remote sensing and {{< video https://www.youtube.com/watch?v=OK_HI3sjbtI >}} +::: {.callout-caution} +## Exercise + +2. Why does the "green belt" in Africa move north and south over the seasons? + +::: + + Leaf phenology is either controlled by temperatures and light availability in temperate, boreal, and tundra ecosystems, or by water availability in warm, seasonally dry climates elsewhere. The water control on vegetation and leaf phenology will be discussed in @sec-ecohydrology. In winter-cold climates (temperate, boreal, and tundra climate zones C, D, and E, see @fig-table_beck), leaf phenology of deciduous plants is driven by the seasonal fluctuations of temperature and radiation. Leaves are shed before winter to avoid damage by cold temperatures and because the limited gains by C assimulation during dark and cold winter months does not outweigh the costs for maintenance of vital functions. In seasonally dry climates (mainly climate zones Am, Aw, see @fig-table_beck), leaves are shed to avoid water loss and avoid dangerous desiccation of the plant. Leaf area also varies more gradually in response to seasonal variations in climate. LAI can have a seasonal cycle also in evergreen forests and shrublands. In grasslands, LAI typically has a strong seasonal cycle, with rapid foliage development and senescence during the typically relatively short vegetation period - but without the typical budbreak, leaf unfolding, and leaf shedding as in deciduous trees. @@ -395,6 +411,13 @@ knitr::include_graphics("images/phenology_patterns.png") +::: {.callout-caution} +## Exercise + +3. With rising temperature under anthropogenic climate change, how do you expect the risk of frost events causing damage on unfolded leaves to change? + +::: + ## Temporal variations of ecosystem C fluxes {#sec-flux-variability} Solar radiation is the "engine" of photosynthesis and thus of the terrestrial carbon cycle. This is expressed by the light use efficiency model of GPP (@eq-luemodel). The temporal and spatial patterns of photosynthetic CO~2~ uptake are therefore driven by the temporal and spatial patterns in solar radiation. The Earth's revolution around its own axis within a day, and its revolution around the sun within a year give rise to *diurnal* and *seasonal* variations in solar radiation, climate, C fluxes, and biological activity in all its forms. (It's a fascinating thought that even physiological and psychological circadian rhythms like wake and sleep patterns are a consequence of solar geometry.) The variations of solar radiation over the course of a day and a year are expressed very differently along different latitudinal positions. @@ -421,7 +444,6 @@ In absence of disturbances influencing ecosystem C fluxes, the net ecosystem exc - During nighttime hours, NEE is relatively stable. During that time, GPP is zero. Hence, *R*~eco~ is relatively stable during the night. - The daily total NEE (@fig-diurnal_cycle c) is smaller than daily totals in GPP and *R*~eco~. At the tropical site, NEE is very small in comparison to GPP and *R*~eco~. This indicates that over the course of one day, the tropical forest site is nearly C neutral. A substantial net C uptake is evident for DE-Hai. This is the C balance for a day in mid July. In general, during summer, temperate, boreal, and tundra ecosystems are a net C sink, and a net C source during winter (see @sec-seasonal-cycle). - ```{r echo=FALSE} #| label: fig-diurnal_cycle #| fig-cap: "Mean diurnal (daily) cycle at northern hemispheric mid-summer (15 July) of (a) incident (top-of-canopy) shortwave radiation, (b) gross primary production, and (d) net ecosystem exchange for four different sites - a temperate deciduous forest (DE-Hai), a boreal needle-leaved forest (FI-Hyy), an arctic heathland in the tundra (US-ICh), and an evergreen moist tropical forest (BR-Sa3). Panel (c) shows the daily total GPP, *R*~eco~, and net ecosystem exchange (NEE). NEE is largely equivalent to the negative of NEP as defined in @sec-nep. Site locations are shown in @fig-biomes_sites_map." @@ -458,6 +480,15 @@ As for the diurnal cycle, also for the seasonal cycle, the pattern in NEE strong # plot created with analysis/plot_gpp_sites_biomes.R knitr::include_graphics("images/seasonal_cycle.png") ``` +::: {.callout-caution} +## Exercise + +4. At what latitude do you expect the largest diurnal cycle amplitude in top-of-atmosphere solar radiation? +5. At what latitude do you expect the largest seasonal amplitude in top-of-atmosphere solar radiation? +6. Where do you expect the smallest seasonal variation in the daily net ecosystem productivity. +7. Consider mature forest ecosystems (*sensu* @odum69sci, @fig-odum) in a moist tropical, temperate and boreal region. Where do you expect the largest annual NEP? Where do you expect the largest daily NEP during the peak growing season? +::: + #### The breathing of the Earth @@ -492,7 +523,7 @@ We start by adopting the hypothetical perspective from the top of the atmosphere Next, we consider the attenuation of solar radiation by the atmosphere, considering cloud cover and the elevation of the land surface. This considers @eq-sin and reflects the geographic pattern of distribution of solar radiation, incident at the land surface (@fig-solar_radiation). Note that $\varphi_0$ and $\tilde{a}$ are considered here to be global constants. Atmospheric attenuation of photysnthetically active solar radiation reduces terrestrial GPP by over 50% (from 2960 to 1442 PgC yr^-1^, @wang14bg). -Not all solar radiation that reaches the land surface is absorbed by active green vegetation. The highest radiation intensities coincide with sparsely vegetated land, including major desert regions. The effect of reduced vegetation cover on GPP is quantified by the factor fAPAR which accounts for the fraction of *absorbed* photosynthetically active radiation (@sec-light-absorption) and reflects water and other limitations on vegetation. Limited vegetation cover and light absorption reduces terrestrial GPP further by over 75% (from 1442 to 322 PgC yr^-1^, @wang14bg). +Not all solar radiation that reaches the land surface is absorbed by active green vegetation. The highest radiation intensities coincide with sparsely vegetated land, including major desert regions. The effect of reduced vegetation cover on GPP is quantified by the factor fAPAR which accounts for the fraction of *absorbed* photosynthetically active radiation (@sec-light-absorption). Limited vegetation cover and light absorption reduces terrestrial GPP further by over 75% (from 1442 to 322 PgC yr^-1^, @wang14bg). At the global scale, temperature has a relatively weak effect on GPP. Considering that photosynthesis is inhibited by freezing temperatures (below 0°C) does not substantially reduce total terrestrial GPP (from 322 to 300 PgC yr^-1^, corresponding to a 7% reduction, @wang14bg). This is because low temperatures coincide with low PAR (or $I_0$) and . Situations of high light and low temperatures are relatively rare. @@ -505,6 +536,14 @@ Finally, physiological effects of CO~2~, temperature, vapour pressure deficit, a knitr::include_graphics("images/gpp_constraints_wang14.png") ``` +::: {.callout-caution} +## Exercise + +8. Which of the factors described above implies the strongest reduction in GPP? What resource limitation may be responsible for this large unrealised C assimilation? + +::: + + ### Global distribution of biomass carbon As explained in @sec-carboncycle_ecosystem, photosynthesis is the ultimate source of all C cycling in ecosystems and accumulating in different pools along the C cascade - from live vegetation to litter and soil organic matter. Variations in GPP thus affect all "downstream" fluxes and pools along the C cascade. A reflection of this is the similarity in the global patterns of GPP and aboveground biomass C (@fig-agc2). In @sec-carboncycle_ecosystem, we considered the C cascade and simulated the dynamics of all pools and fluxes in response to a change in GPP. We showed that, assuming that allocation, SOM partitioning factors, and pool-specific turnover times are constant, GPP linearly scales all downstream fluxes and pools. We also noted that in reality, allocation, SOM partitioning, and turnover times cannot be assumed constant over time. When considering variations across space - across different climates, soils, and ecosystem types - constancy in these factors is probably an even less justified assumption. @@ -622,6 +661,16 @@ As mentioned above, peatlands store 500-700 PgC (more than all biomass C globall Peatlands, as all wetlands, are also a major source of methane (CH~4~). Under anoxic soil conditions, CH~4~ is produced, rather than C being oxidized and CO~2~ released (@sec-ghg). +::: {.callout-caution} +## Exercise + +9. Water table variations affect the C balance of a peatland. If it falls, how do you expect NPP, *R*~eco~, and NEP of a peatland, and the turnover time of C in soil organic matter to change? +10. What other greenhouse gas may be influenced by a declining water table in a peatland? In what direction would its emission change? +11. Why are soils dark in the Seeland of the Canton of Bern? +12. Why are wooden artefacts of the Bronze age preserved in some places but not in others? Do you know any such wooden artefacts? + +::: + ### Permafrost {#sec-permafrost} @@ -646,6 +695,15 @@ The presence of permafrost constrains the volume available for plant rooting, ve With rapid warming in northern high latitudes, permafrost temperatures are rising throughout the Arctic [@biskaborn19natcomm]. Implications of permafrost melting for greenhouse gases (CO~2~ and CH~4~) and the climate-land Earth system feedback are discussed in later chapters. +::: {.callout-caution} +## Exercise + +13. Why does the melting of permafrost affect methane emissions? + +::: + + + diff --git a/book/gpp.qmd b/book/gpp.qmd index 37cf38c..543c21b 100644 --- a/book/gpp.qmd +++ b/book/gpp.qmd @@ -173,6 +173,8 @@ Second, the energy flux of direct radiation is much higher than that of diffuse Light use efficiency (LUE, @eq-luemodel) measures the efficiency at which absorbed photons (radiation) are converted into C in the form of sugars. LUE is an ecosystem-level quantity and reflects the collective photosynthetic CO~2~ uptake efficiency of all leaves in the canopy. This efficiency is determined by the leaf physiology and the activity of the photosynthetic machinery. In this subsection, we dive very deep into the physiology of photosynthesis. + + CO~2~ uptake by the leaf happens by diffusion through pores at the leaf surface (*stomata*, @fig-stomata). The opening of stomata is dynamically regulated in response to environmental conditions (light, CO~2~, temperature, atmospheric humidity) and controls the conductance to CO~2~ diffusion. The CO~2~ uptake flux by diffusive transport through stomata can thus be described by *Fick's law* which states that the flux is proportional to the concentration difference of CO~2~ inside and outside the leaf. $$ A = g_s(c_a - c_i) diff --git a/book/images/canopy_resistance.png b/book/images/canopy_resistance.png new file mode 100644 index 0000000..74cb0ef Binary files /dev/null and b/book/images/canopy_resistance.png differ diff --git a/book/images/diurnal_cycle_radiation.png b/book/images/diurnal_cycle_radiation.png new file mode 100644 index 0000000..7d90de5 Binary files /dev/null and b/book/images/diurnal_cycle_radiation.png differ diff --git a/book/images/energybalance.key b/book/images/energybalance.key new file mode 100755 index 0000000..63e5165 Binary files /dev/null and b/book/images/energybalance.key differ diff --git a/book/images/energybalance.pdf b/book/images/energybalance.pdf new file mode 100644 index 0000000..34d4f50 Binary files /dev/null and b/book/images/energybalance.pdf differ diff --git a/book/images/energybalance.png b/book/images/energybalance.png new file mode 100644 index 0000000..59f4183 Binary files /dev/null and b/book/images/energybalance.png differ diff --git a/book/images/seasonal_cycle_radiation.png b/book/images/seasonal_cycle_radiation.png new file mode 100644 index 0000000..94361d0 Binary files /dev/null and b/book/images/seasonal_cycle_radiation.png differ diff --git a/book/images/snow_masking_bonan.jpg b/book/images/snow_masking_bonan.jpg new file mode 100644 index 0000000..d4f5c85 Binary files /dev/null and b/book/images/snow_masking_bonan.jpg differ diff --git a/book/landclimate.qmd b/book/landclimate.qmd index 10bacab..d3e5d24 100644 --- a/book/landclimate.qmd +++ b/book/landclimate.qmd @@ -1,6 +1,536 @@ +--- +format: + html: + fig-width: 5 + fig-height: 4 + pdf: + fig-width: 7 + fig-height: 5 +--- + # Land-climate interactions {#sec-landclimate} -Coming soon. +In previous chapters, we have investigated the interactions between the atmosphere and the land surface through the exchange of CO~2~. In this, chapter, we focus on interactions through physical processes, including the radiative energy balance and heat fluxes at the land surface. Since these Earth system interactions are also mediated to a large degree through biology, they are referred to as *biogeophysical* processes. This is in contrast to the exchange of CO~2~ and nutrient fluxes, referred to as *biogeochemical* processes. + +The partitioning of heat fluxes into a sensible (hot air) and latent (moist air) heat flux is strongly affected by the availability of water for evaporation and transpiration. All aspects that relate to the water cycle, how it is affected by vegetation, and how it affects the surface energy balance and limits vegetation activity, are introduced in @sec-ecohydrology. However, land-climate interactions are also important in absence of effects by water limitation. The surface properties, determined by the land cover type, its roughness, and (for vegetation) its conduciveness to evaporating water, have a strong influence on the atmosphere and climatic conditions, especially near the. surface - where life happens. Climate and numerical weather prediction models require the fluxes of energy, momentum, and water vapour for simulating atmospheric processes. This chapter introduces how properties of the vegetated land surface affect these fluxes. + +## Components of surface radiation + +Analogous to solar radiation being the "engine" of biogeochemical processes through its control on photosynthesis, net radiation is the engine of energy and water vapour fluxes between the atmosphere and the land surface. Net radiation ($R_n$) is the net of incoming (or incident) shortwave radiation ($S\downarrow$) minus outgoing shortwave (solar) radiation ($S\uparrow$) and longwave radiation ($L\downarrow$ and $L\uparrow$) at the land surface (@fig-netrad a). +$$ +R_n = S\downarrow - S\uparrow + L\downarrow - L\uparrow +$$ {#eq-netrad} + +```{r echo=FALSE} +#| label: fig-netrad +#| fig-cap: "Land surface energy partitioning. (a) Components of net radiation. (b) Energy balance of the land surface." +#| out-width: 60% +knitr::include_graphics("images/energybalance.png") +``` + +### Incident shortwave radiation + +Incident shortwave (solar) radiation $S\downarrow$ was referred to in @sec-global-patterns as $I_\mathrm{0}$. (Nomenclature will be homogenized across chapters in future issues this book.) It varies over the course of a day and a year, and is affected by the presence of clouds and the altitude of the land surface above sea level, as described in @sec-global-patterns. As shown in @fig-surface-energy-balance, the presence of clouds is a strong determinant of the incoming solar radiation energy flux, when considering global means (including the sea surface) under actual conditions and under cloud-free conditions (160 vs. 214 W^-2^). + +```{r echo=FALSE} +#| label: fig-surface-energy-balance +#| fig-cap: "Schematic representation of the global mean energy budget of the Earth (upper panel), and its equivalent without considerations of cloud effects (lower panel). Numbers indicate best estimates for the magnitudes of the globally averaged energy balance components in W^-2^ together with their uncertainty ranges in parentheses (5–95% confidence range), representing climate conditions at the beginning of the 21st century. Note that the cloud-free energy budget shown in the lower panel is not the one that Earth would achieve in equilibrium when no clouds could form. It rather represents the global mean fluxes as determined solely by removing the clouds but otherwise retaining the entire atmospheric structure. This enables the quantification of the effects of clouds on the Earth energy budget and corresponds to the way clear-sky fluxes are calculated in climate models. Thus, the cloud-free energy budget is not closed and therefore the sensible and latent heat fluxes are not quantified in the lower panel. Figure and caption from @IPCC_2021_WGI_Ch_7." +#| out-width: 100% +knitr::include_graphics("https://www.ipcc.ch/report/ar6/wg1/downloads/figures/IPCC_AR6_WGI_Figure_7_2.png") +``` +### Outgoing shortwave radiation and albedo + +Outgoing shortwave radiation at the land surface ($S\uparrow$) is determined by the albedo ($\alpha$) - the fraction of the incident radiation that gets reflected. +$$ +S\uparrow = \alpha S\downarrow +$$ {#eq-albedo} +The albedo varies across different vegetation and other types of Earth surface covers. Values range from 0.8-0.95 for fresh snow, to 0.2-0.45 for a desert surface, 0.05–0.40 for bare soil, 0.05-0.26 for vegetation, and 0.03–0.10 for water [@oke87]. Variations across different vegetation types are substantial as shown also in @fig-albedo-vegtypes. This indicates that the amount of outgoing shortwave radiation, and with it net radiation, is strongly affected by vegetation types. + +```{r warning=FALSE, message=FALSE} +#| code-fold: true +#| label: fig-albedo-vegtypes +#| fig-cap: "Albedo values for different vegetation types. Each point represents a site where surface radiation fluxes were measured. Vegetation types are described in @tbl-vegtype. Data from @cescatti12rse." +#| out-width: 60% + +library(dplyr) +library(readr) +library(ggplot2) + +df <- read_csv(here::here("data-raw/albedo_cescatti.csv")) + +df |> + ggplot(aes( + x = reorder(vegtype, albedo_insitu, decreasing = TRUE), + y = albedo_insitu + )) + + geom_boxplot(fill = "grey70") + + geom_jitter(width = 0.15) + + theme_classic() + + labs( + x = "Vegetation type", + y = "Albedo" + ) +``` + +The diurnal variation of albedo is U-shaped - high at sunrise and sunset, and lower in between. Over the seasons, land surface albedo is affected by variations in LAI and by the presence of snow. During winter and in the presence of snow, the albedo of a grassland or a cropland can reach very high values - determined by the albedo of snow. In contrast, the albedo of a forest, even in the presence of snow, remains. much lower (0.2-0.4, @zhao14ecolmono). This is because trees are not fully "submerged" in the snow pack, as opposed to grasses, and the exposed canopy absorbs a substantial fraction of the incident radiation. This effect of forest vs. non-forest covers on the surface energy balance during winter is referred to as the *snow masking effect*. Climate model simulations suggest that the low surface albedo during winter warms climate compared compared to a snow-covered grassland. Thus, deforestation/afforestation of boreal forest has the greatest biogeophysical effect of all biomes on temperatures [@bonan08sci]. + +```{r echo=FALSE} +#| label: fig-snow-masking +#| fig-cap: "Snow-free vs. snow-covered surface albedo for different vegetation types. Figure from @bonan08sci." +#| out-width: 60% +knitr::include_graphics("images/snow_masking_bonan.jpg") +``` + +Note that albedo also depends on the wavelength of the radiation, on the solar zenith angle (lower at noon), and on whether the solar radiation is diffuse or direct. Thus, albedo varies over time, $\alpha$ in @eq-albedo represents the albedo for shortwave radiation, and values shown in @fig-albedo-vegtypes are averages across the seasons. + + +### Longwave radiation + +Every material body with a temperature above the absolute zero (0 K $= -$ 273.15°C) emits radiative energy. The total radiative energy flux (integrated across wavelengths) and the wavelengths at which radiation is emitted depend on the temperature of the body and its emissivity. The surface of the sun is about 6000 K and emits radiation in the shorwave spectrum which is visible to the human eye. The Earth surface is on average 288 K and emits radiation in the longwave spectrum which is not visible to the eye. +```{r warning=FALSE, message=FALSE, fig.height=3, fig.width=8} +#| code-fold: true +#| label: fig-wavelengths +#| fig-cap: "Spectral distribution of the radiative energy flux (a) for a body of 6000 K (the sun) and (b) a body of 288 K (the earth), assuming an emissivity of 1. *Emittance* (the y-axis) is the density of the radiative energy flux per unit wavelength spectrum. Note the different magnitudes of the emittances and wavelengths for the two bodies. Example following @bonan." +#| out-width: 100% + +library(ggplot2) +library(cowplot) + +# Formula for Planck's Law +calc_planck <- function(lambda, temp){ + h <- 6.626e-34 # Planck’s constant + c <- 3e8 # speed of light + k <- 1.38e-23 # Stefan Boltzmann constant + out <- (2*pi*h*c^2) / (lambda^5*(exp(h*c/(k*lambda*temp)) - 1)) + return(out) +} + +gg1 <- ggplot() + + geom_function( + fun = calc_planck, + args = list( + temp = 6000 + ) + ) + + xlim(0, 0.4e-5) + + theme_classic() + + labs( + title = "Sun", + subtitle = "6000 K", + x = "Wavelength (m)", + y = expression(paste("Emittance (W m"^-2, " m"^-1, ")")) + ) + +gg2 <- ggplot() + + geom_function( + fun = calc_planck, + args = list( + temp = 288 + ) + ) + + xlim(0, 1e-4) + + theme_classic() + + labs( + title = "Earth", + subtitle = "288 K", + x = "Wavelength (m)", + y = expression(paste("Emittance (W m"^-2, " m"^-1, ")")) + ) + +plot_grid( + gg1, + gg2, + labels = c("a", "b") +) +``` +According to the Stefan-Boltzmann law, the total radiative energy flux scales with the temperature of the body $T$ and an emissivity $\varepsilon$: +$$ +R = \varepsilon \sigma T^4 +$$ {#eq-boltzmann} +$\sigma$ is the Stefan Boltzmann constant and has a value of $5.670\times 10^{-8}$ W m^-2^ K^-4^. Here, $T$ is expressed in Kelvin. $R$ is the integral under the curve in @fig-wavelengths. $R$ stands for either longwave or shortwave radiation, as referred to as $L$ and $S$ in @eq-netrad. + +The longwave radiation components are a substantial fraction of net radiation. In @fig-surface-energy-balance, they are referred to as "thermal". The outgoing longwave radiation $L\uparrow$ is on average 398 W m^-2^ and is a function of the Earth's surface temperature and emissivity, following the Stefan-Boltzmann law (@eq-boltzmann), plus the incident longwave radiation reflected (not absorbed) by the surface: +$$ +L\uparrow = \varepsilon \sigma T^4 + (1-\varepsilon) L\downarrow +$$ {#eq-longwave-up} +Note that $\varepsilon$ is used in @eq-longwave-up as the absorbtivity and in @eq-boltzmann as the emissivity. This is because the two are equal. + +Clouds absorb radiation emitted by the earth surface and re-emit a large fraction of it. In absence of clouds, much less radiation is absorbed and re-emitted by the atmosphere. Therefore, the incoming longwave radiation depends strongly on the presence of clouds. + +::: {.callout-note} + +### Net radiation + +Using @eq-albedo and @eq-longwave-up, the net radiation at the earth surface (@eq-netrad) can be expressed as +$$ +R_n = (1-\alpha) S\downarrow + \varepsilon L\downarrow - \varepsilon \sigma T^4 +$$ +::: + + + +::: {.callout-caution} +## Exercise + +1. Calculate the albedo from 'all sky' conditions in @fig-surface-energy-balance. +::: + +## Surface energy balance + +The energy available from radiative fluxes is converted into a sensible heat flux ($H$), a latent heat flux ($\lambda E$), and a ground heat flux ($G$, @fig-netrad b). Following energy conservation, net radiation is equal to the sum of these three heat fluxes. +$$ +R_n = H + \lambda E + G +$$ {#eq-energybalance} + +The components of the surface energy balance (@eq-energybalance) are commonly expressed in energy units (W m^-2^). Sensible heat is determined by the temperature of air. Latent heat is the energy contained by evaporated water. The latent heat flux ($\lambda E$) can also be expressed as a mass flux of water vapour ($E$). $E$ is the mass flux of water vapor, e.g., expressed in units of g m^-2^ s^-1^. $\lambda$ is the latent heat of vaporization and converts the mass units into energy units. It measures how much energy (Joules, J) is needed vapourize 1 g of water at constant temperature. $\lambda$ is 2.466 MJ kg^-1^ at 15°C and has a slight dependence on temperature, decreasing linearly by about 3% from an air temperature of 0°C to 30°C (see code below using the {bigleaf} R package [@knauer18plosone]). + +```{r warning=FALSE, message=FALSE} +#| code-fold: false +library(bigleaf) # This package contains lots of useful functions and parameters for land-atmosphere exchange. Reference: Knauer et al. 2018 https://dx.plos.org/10.1371/journal.pone.0201114 +latent.heat.vaporization(30) / latent.heat.vaporization(0) +``` + +The sensible and latent heat fluxes are transported vertically, away from or to the land surface through convective transport. That is, through turbulences that mix the air and lead to a net vertical transport of heat and water vapor. Whether net fluxes are pointed upwards or downwards depends on the sign of the net radiation (see fig xxx) and typically changes between night and daytime. Note that the latent heat flux can be negative, leading to a net flux of water vapor towards the surface. The respective water mass *condensates* at the surface (of leaves) and can be a considerable fraction of the ecosystem water balance. The ground heat flux $G$ buffers variations of net radiation, absorbing energy and removing heat from the surface during the day and summer and releasing heat during night and winter. + +### Energy partitioning + +Physical and biological properties of the land surface determine not only the net radiation through (mainly) effects of the albedo, but also strongly influence the *partitioning* of net radiation into sensible and latent heat fluxes in @eq-energybalance. Thus, the spatial distribution of vegetation and other land cover types, and the response of plants to temporal changes in the environment (and thus on LAI and stomatal conductance) influence the energy partitioning at the land surface and near-surface atmospheric conditions. In other words, we have to understand vegetation and its response to the environment in order to understand land-climate interactions and near-surface climate. + +The key properties influencing the partitioning into the sensible and latent heat flux are the aerodynamic conductance to heat transfer ($G_\mathrm{ah}$) and the surface conductance to water vapor transport ($G_\mathrm{sw}$). How they influence the latent heat flux is shown in @fig-penmanmonteith. The aerodynamic conductance depends on the roughness of the surface and on wind speed. Taller vegetation has a higher roughness and a higher aerodynamic conductance. Roughness also increases with LAI. + +On vegetated surfaces, the surface conductance to water vapor transport is strongly influenced by the stomatal conductance ($g_s$, @eq-photo-fick and @eq-transpiration) and by the LAI. Water evaporation from rock, soil, leaf, or branch surfaces contributes to surface conductance and occurs also from non-vegetated surfaces. In a closed canopy, surface conductance is dominated by the signal by stomatal conductance. When leaves are active and photosynthesizing (high light), and when water stress is low (high soil moisture, low VPD), stomatal conductance is high. How water availability and vegetation regulate stomatal conductance and thus the latent heat flux and energy partitioning at the land surface is introduced in @sec-ecohydrology. + +```{r echo=FALSE} +#| label: fig-surface-resistance +#| fig-cap: "Surface resistance and normalized latent heat flux for different vegetation types. Normalization removes effects of different net radiation on latent heat fluxes. Canopy resistance is equivalent to the inverse of surface conductance described in this chapter. Figure from @bonan08sci." +#| out-width: 70% +knitr::include_graphics("images/canopy_resistance.png") +``` + +::: {.callout-note} + +### Dependency of energy partitioning on conductances + +Given the atmospheric condition (net radiation $R_n$ and the vapor pressure deficit of the ambient air $D_a$), the latent heat flux can be modelled following the Penman-Monteith equation: +$$ +\lambda E = \frac{s(R_n - G) + \rho c_p D_a G_\mathrm{ah}}{s + \gamma (1 + G_\mathrm{ah}/G_\mathrm{sw})} +$$ {#eq-pm} + +- $s$ is the slope of the saturation vapor pressure versus temperature curve (kPa K^−1^) +- $\rho$ is the density of air (kg m^-3^) +- $c_p$ is the heat capacity of dry air (J K^−1^ kg^−1^) +- $\gamma$ is the psychrometric constant (kPa K^−1^) + +A derivation of the Penman-Monteith equation is given by @bonan, Ch. 12.7. + +With this, and given the atmospheric condition, we can visualise the dependency of the latent heat flux on the aerodynamic and the surface conductances. Let's assume $R_n =$ 400 W m^-1^. + +```{r warning=FALSE, message=FALSE, fig.height=3, fig.width=8} +#| code-fold: true +#| label: fig-penmanmonteith +#| fig-cap: "Dependency of the latent heat flux on (a) the aerodynamic conductance and (b) the surface conductance following the Penman-Monteith equation (@eq-pm)." +#| out-width: 100% + +library(bigleaf) + +# calculate latent heat flux using penman monteith +calc_le_pm <- function(netrad, vpd, temp, g_ah, g_sw){ + # ARGUMENTS + # netrad: net radiation + # vpd: vapour pressure deficit + # temp: ambient air temperature + # patm: atmospheric pressure (kPa) + # aerodynamic conductance to heat transport, in mass units (m s-1) + # surface conductance to water vapor transport, in mass units (m s-1) + + # using standard atmospheric pressure + patm <- bigleaf.constants()$pressure0 * 1e-3 + + # slope of the saturation vapor pressure, using "Sonntag_1990" in Bigleaf + s <- Esat.slope(temp)$Delta + + # density of air + rho <- air.density(temp, patm) + + # heat capacity of dry air + cp <- bigleaf.constants()$cp + + # psychrometric constant + gamma <- psychrometric.constant(temp, patm) + + # assuming G = 0; conductances are in mass units + out <- (s * netrad + rho * cp * vpd * g_ah) / (s + gamma * (1 + g_ah / g_sw)) + + return(out) +} + +# # test our function +# calc_le_pm( +# netrad = 100, +# vpd = 1, +# temp = 15, +# g_ah = 0.1, # mass units (m s-1) +# g_sw = mol.to.ms(0.6, Tair = temp, pressure = patm, constants = bigleaf.constants()) # convert molar units (mol m-2 s-1) to mass units +# ) +# +# # test built-in function +# potential.ET( +# Gs_pot = 0.6, # molar units (mol m-2 s-1) +# Tair = 15, +# pressure = bigleaf.constants()$pressure0 * 1e-3, +# VPD = 1, +# Ga = 0.1, # mass units +# Rn = 100, +# approach = "Penman-Monteith" +# )[,"LE_pot"] + +# plot vs aerodynamic conductance +df <- expand.grid( + g_ah = seq(0, 0.3, by = 0.01), + vpd = seq(0, 5, by = 0.5) +) |> + as_tibble() |> + mutate( + le = purrr::map2_dbl( + g_ah, + vpd, + ~calc_le_pm( + netrad = 400, + vpd = .y, + temp = 15, + g_ah = .x, + g_sw = mol.to.ms( + 0.6, + Tair = 15, + pressure = bigleaf.constants()$pressure0 * 1e-3, + constants = bigleaf.constants() + )) + ) + ) + +gg1 <- df |> + ggplot() + + geom_line(aes(g_ah, le, color = vpd, group = vpd)) + + scale_color_viridis_c(name = "VPD (kPa)") + + geom_hline(yintercept = 0, linetype = "dotted") + + theme_classic() + + labs( + x = expression(paste("Aerodynamic conductance (m s"^-1, ")")), + y = expression(paste(lambda, italic("E"), " (W m"^-2, ")")) + ) + + theme(legend.position="none") + +# plot vs surface conductance +df <- expand.grid( + g_sw = seq(0, 0.2, by = 0.01), + vpd = seq(0, 5, by = 0.5) +) |> + as_tibble() |> + mutate( + le = purrr::map2_dbl( + g_sw, + vpd, + ~calc_le_pm( + netrad = 400, + vpd = .y, + temp = 15, + g_ah = 0.1, + g_sw = .x) + ) + ) + +gg2 <- df |> + ggplot() + + geom_line(aes(g_sw, le, color = vpd, group = vpd)) + + scale_color_viridis_c(name = "VPD (kPa)") + + theme_classic() + + labs( + x = expression(paste("Surface conductance (m s"^-1, ")")), + y = expression(paste(lambda, italic("E"), " (W m"^-2, ")")) + ) + +plot_grid( + gg1, + gg2, + labels = c("a", "b"), + rel_widths = c(0.77, 1) +) +``` + +@fig-penmanmonteith shows an interesting interactive effect of aerodynamic conductance and the VPD of ambient air. At low VPD, plant transpiration is reduced, thus limiting the latent heat flux ($\lambda E$). Under such conditions, the $\lambda E$ declines with increasing aerodynamic conductance. This is because $H$ rises faster than $\lambda E$ with increasing aerodynamic conductance and consumes a larger share of the net radiation. Under moderate-to-high VPD (above $\sim$1 kPa in this example), $\lambda E$ rises with increasing aerodynamic conductance. + +Under conditions of very low aerodynamic conductance (e.g., in a short-statured grassland that has a low surface roughness and under stable atmospheric conditions, e.g., during an inversion), the latent heat flux does not approach zero. This is because as long as there is positive net radiation (here 400 W m^-2^), the surface (skin temperature) heats up relative to the ambient air and creates a positive VPD at the leaf surface (even if VPD of ambient air is zero). These aspects drive a continued positive $H$ and $\lambda E$ and the rise of the surface boundary layer. + + +::: + +### Evaporative fraction and Bowen ratio + +When considering totals over longer periods of time (over at least one annual cycle), the ground heat flux can be neglected since the ground doesn't gradually heat up over time. Hence, the sum of the latent and sensible heat fluxes add up to match net radiation. The *evaporative fraction* quantifies the fraction of available energy from net radiation goes into the latent heat flux. +$$ +\mathrm{EF} = \lambda E/R_n +$$ +The remainder goes into the sensible heat flux. The Bowen ratio expresses the same relation and is defined as +$$ +B = H/\lambda E +$$ +Quantifications of EF or $B$ give important insights beyond the "first-order control" by net radiation. Variations across space reflect surface properties and vegetation activity. In forests, the proportion of evapotranspiration to available energy is typically lower compared to certain crops, and it's even lower in conifer forests than in deciduous broadleaf forests (@fig-surface-resistance). + +## Energy fluxes across biomes + +As for C fluxes in @sec-global-patterns, we turn to investigating diurnal and seasonal patterns in fluxes - this time the energy fluxes. The patterns observed at the same sites as used before illustrate general differences across biomes and how different characteristics of the vegetation cover affect the surface energy balance and flux partitioning. In addition to the temperate broadleaved forest site (Hainich Forest, DE-Hai), we consider a coniferous forest (DE-Tha) and a grassland site (DE-Gri) that are located closeby (and thus experience largely identical meterological conditions, ignoring feedbacks from vegetation-atmosphere fluxes that affect near-surface climate). + +```{r, message=FALSE, warning=FALSE} +#| code-fold: true +#| label: fig-site-info +#| fig-cap: "Site meta information. Vegetation types are described in @tbl-vegtype. Climate zones are described in @fig-table_beck." + +df_sites <- read_csv(here::here("data/fdk_site_info.csv")) + +# chose representative sites +use_sites <- c( + "FI-Hyy", # Boreal Forests/Taiga + "DE-Hai", # Temperate Broadleaf & Mixed Forests + "DE-Tha", # Temperate Coniferous + "DE-Gri", # Grassland just next to DE-Tha + "BR-Sa3", # Tropical + "US-ICh" # Tundra +) + +# subset data +df_sites |> + filter(sitename %in% use_sites) |> + mutate( + lon = format(lon, digits = 3), + lat = format(lat, digits = 3), + elv = format(elv, digits = 3) + ) |> + select( + Site = sitename, + Longitude = lon, + Latitude = lat, + Elevation = elv, + `Vegetation type` = igbp_land_use, + `Climate zone` = koeppen_code + ) |> + knitr::kable() +``` + +### Diurnal variations + + + +```{r echo=FALSE} +#| label: fig-energy-diurnal +#| fig-cap: "Diurnal variations of energy fluxes at sites representing different biomes, taken as the average for each half-hour of a day in July." +#| out-width: 100% +# figure created with analysis/plot_radiation_sites_biomes.R +knitr::include_graphics("images/diurnal_cycle_radiation.png") +``` + +::: {.callout-caution} +## Exercise + +The diurnal course of energy fluxes is considered for an average day in July for each site. Describe key patterns and differences between sites and relate them to your knowledge about radiation, energy partitioning, and land surface and vegetation properties. + +2. How does net radiation compare across sites in the top row of @fig-energy-diurnal? What component of the net radiation (@fig-netrad a) do you think is responsible for the observed differences? What surface property may contribute to the differences in net radiation and does it explain why the tropical site has a higher mid-day net radiation peak than the boreal site? +3. How does net radiation compare across the three sites in the temperate biome? Can vegetation properties that affect the reflectance explain why the mid-day peak net radiation is lower for DE-Gri han for DE-Tha? +4. During night-time, net radiation is negative at all sites. Which of the four components of net radiation is largest during nighttime? +5. Is the land surface losing or gaining energy during the night? What flux drives this gain/loss ($H$ or $\lambda E$)? +6. At what site is the nighttime energy gain/loss highest? What surface property may be responsible for this? +7. Chararcterise the Bowen ratio at mid-day for all sites (greater than, close to, or smaller than 1). +8. What site has the highest mid-day Bowen ratio? What site has the lowest mid-day Bowen ratio? +9. Assuming all sites had the same aerodynamic and surface conductances, which site may have the highest mid-day VPD? Is this a valid assumption? If not, how do you expect conductances to differ among sites? +10. The temperate coniferous forest site (which one?) has the higher mid-day net radiation than the temperate broadleaved fores site (which one?). Yet, its latent heat flux is lower than for the latter. Assuming they had the same VPD and aerodynamic conductance, what other surface property may explain this difference? +11. The temperate grassland has a higher latent heat flux than the two temperate forest sites. Can the difference in vegetation height explain the difference in the latent heat flux? +12. What site has the highest sensible heat flux? If you were a paraglider, over which site do you expect to find the strongest up-lift to take you to your next destination? + +::: + + + + + + + + + +### Seasonal variations + +```{r echo=FALSE} +#| label: fig-energy-seasonal +#| fig-cap: "Seasonal variations of energy fluxes at sites representing different biomes." +#| out-width: 100% +# figure created with analysis/plot_radiation_sites_biomes.R +knitr::include_graphics("images/seasonal_cycle_radiation.png") +``` + +::: {.callout-caution} +## Exercise + +12. Which site has the highest net radiation around May? What surface property may this be related to? +13. Which site has the lowest summertime net radiation? How does this compare to differences in summertime solar radiation among these sites? What other components of net radiation do you expect to explain the low summertime net radiation at this site? +14. What could explain the different seasonal course of the latent and sensible heat flux at DE-Hai, while they vary in parallel at DE-Tha? +15. What may be the cause of the delayed springtime increase of $\lambda E$, compared to $H$ at FI-Hyy? +16. What site loses the largest amount of water on an annual basis? +17. Do you find other interesting patterns and interpretable differences between sites? + +::: + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/book/references.bib b/book/references.bib index 2cd8749..e58a7d4 100644 --- a/book/references.bib +++ b/book/references.bib @@ -1992,3 +1992,87 @@ @article{biskaborn19natcomm file = {Full Text PDF:/Users/benjaminstocker/Zotero/storage/H7ZM3HVG/Biskaborn et al. - 2019 - Permafrost is warming at a global scale.pdf:application/pdf}, } +# Downloaded from https://github.com/openclimatedata/ipcc-bibtex/blob/main/ar6-wg-i.bib +@incollection{IPCC_2021_WGI_Ch_7, + address = {Cambridge, UK and New York, NY, USA}, + author = {Forster, P. and Storelvmo, T. and Armour, K. and Collins, W. and Dufresne, J.-L. and Frame, D. and Lunt, D.J. and Mauritsen, T. and Palmer, M.D. and Watanabe, M. and Wild, M. and Zhang, H.}, + booktitle = {Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change}, + chapter = {7}, + doi = {10.1017/9781009157896.009}, + editor = {Masson-Delmotte, V. and Zhai, P. and Pirani, A. and Connors, S. L. and Péan, C. and Berger, S. and Caud, N. and Chen, Y. and Goldfarb, L. and Gomis, M. I. and Huang, M. and Leitzell, K. and Lonnoy, E. and Matthews, J. B. R. and Maycock, T. K. and Waterfield, T. and Yelekçi, O. and Yu, R. and Zhou, B.}, + publisher = {Cambridge University Press}, + title = {The Earth's Energy Budget, Climate Feedbacks, and Climate Sensitivity}, + type = {Book Section}, + url = {https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6_WGI_Chapter07.pdf}, + year = {2021} +} + + +@article{cescatti12rse, + title = {Intercomparison of {MODIS} albedo retrievals and in situ measurements across the global {FLUXNET} network}, + volume = {121}, + issn = {0034-4257}, + url = {https://www.sciencedirect.com/science/article/pii/S0034425712001125}, + doi = {10.1016/j.rse.2012.02.019}, + abstract = {Surface albedo is a key parameter in the Earth's energy balance since it affects the amount of solar radiation directly absorbed at the planet surface. Its variability in time and space can be globally retrieved through the use of remote sensing products. To evaluate and improve the quality of satellite retrievals, careful intercomparisons with in situ measurements of surface albedo are crucial. For this purpose we compared MODIS albedo retrievals with surface measurements taken at 53 FLUXNET sites that met strict conditions of land cover homogeneity. A good agreement between mean yearly values of satellite retrievals and in situ measurements was found (r2=0.82). The mismatch is correlated with the spatial heterogeneity of surface albedo, stressing the relevance of land cover homogeneity when comparing point to pixel data. When the seasonal patterns of MODIS albedo are considered for different plant functional types, the match with surface observations is extremely good at all forest sites. On the contrary, satellite retrievals at non-forested sites (grasslands, savannas, croplands) underestimate in situ measurements across the seasonal cycle. The mismatch observed at grassland and cropland sites is likely due to the extreme fragmentation of these landscapes, as confirmed by geostatistical attributes derived from high resolution scenes.}, + urldate = {2024-04-29}, + journal = {Remote Sensing of Environment}, + author = {Cescatti, Alessandro and Marcolla, Barbara and Santhana Vannan, Suresh K. and Pan, Jerry Yun and Román, Miguel O. and Yang, Xiaoyuan and Ciais, Philippe and Cook, Robert B. and Law, Beverly E. and Matteucci, Giorgio and Migliavacca, Mirco and Moors, Eddy and Richardson, Andrew D. and Seufert, Günther and Schaaf, Crystal B.}, + month = jun, + year = {2012}, + keywords = {FLUXNET, LES, MODIS, Plant functional types, Remote sensing, Surface albedo, Terrestrial ecosystems, Validation}, + pages = {323--334}, + file = {Cescatti et al. - 2012 - Intercomparison of MODIS albedo retrievals and in .pdf:/Users/benjaminstocker/Zotero/storage/IMXJJT3A/Cescatti et al. - 2012 - Intercomparison of MODIS albedo retrievals and in .pdf:application/pdf;ScienceDirect Snapshot:/Users/benjaminstocker/Zotero/storage/QDWUANPT/S0034425712001125.html:text/html}, +} + +@article{zhao14ecolmono, + title = {Biophysical forcings of land-use changes from potential forestry activities in {North} {America}}, + volume = {84}, + issn = {0012-9615}, + url = {https://www.jstor.org/stable/43187893}, + abstract = {Land-use changes through forestry and other activities alter not just carbon storage, but biophysical properties, including albedo, surface roughness, and canopy conductance, all of which affect temperature. This study assessed the biophysical forcings and climatic impact of vegetation replacement across North America by comparing satellitederived albedo, land surface temperature (LST), and evapotranspiration (ET) between adjacent vegetation types. We calculated radiative forcings (RF) for potential local conversions from croplands (CRO) or grasslands (GRA) to evergreen needleleaf (ENF) or deciduous broadleaf (DBF) forests. Forests generally had lower albedo than adjacent grasslands or croplands, particularly in locations with snow. They also had warmer nighttime LST, cooler daily and daytime LST in warm seasons, and smaller daily LST ranges. Darker forest surfaces induced positive RFs, dampening the cooling effect of carbon sequestration. The mean (±SD) albedo-induced RFs for each land conversion were equivalent to carbon emissions of 2.2 ± 0.7 kg C/m² (GRA-ENF), 2.0 ± 0.6 kg C/m² (CRO-ENF), 0.90 ± 0.50 kg C/m² (CRO-DBF), and 0.73 ± 0.22 kg C/m² (GRA-DBF), suggesting that, given the same carbon sequestration potential, a larger net cooling (integrated globally) is expected for planting DBF than ENF. Both changes in LST and ET induce longwave RFs that sometimes had values comparable to or even larger than albedo-induced shortwave RFs. Sensible heat flux, on average, increased when replacing CRO with ENF, but decreased for conversions to DBF, suggesting that DBF tends to cool near-surface air locally while ENF tends to warm it. This local temperature effect showed some seasonal variation and spatial dependence, but did not differ strongly by latitude. Overall, our results show that a carbon-centric accounting is, in . many cases, insufficient for climate mitigation policies. Where afforestation or reforestation occurs, however, deciduous broadleaf trees are likely to produce stronger cooling benefits than evergreen needleleaf trees provide.}, + number = {2}, + urldate = {2024-04-29}, + journal = {Ecological Monographs}, + author = {Zhao, Kaiguang and Jackson, Robert B.}, + year = {2014}, + note = {Publisher: [Wiley, Ecological Society of America]}, + keywords = {albedo, LES}, + pages = {329--353}, + file = {JSTOR Full Text PDF:/Users/benjaminstocker/Zotero/storage/IJN5L6PJ/Zhao and Jackson - 2014 - Biophysical forcings of land-use changes from pote.pdf:application/pdf}, +} + +@book{oke87, + address = {London}, + edition = {2}, + title = {Boundary {Layer} {Climates}}, + isbn = {978-0-203-40721-9}, + abstract = {This modern climatology textbook explains those climates formed near the ground in terms of the cycling of energy and mass through systems.}, + publisher = {Routledge}, + author = {Oke, T. R.}, + month = dec, + year = {1987}, + doi = {10.4324/9780203407219}, + keywords = {LES}, +} + +@article{knauer18plosone, + title = {Bigleaf—{An} {R} package for the calculation of physical and physiological ecosystem properties from eddy covariance data}, + volume = {13}, + issn = {1932-6203}, + url = {https://dx.plos.org/10.1371/journal.pone.0201114}, + doi = {10.1371/journal.pone.0201114}, + abstract = {We present the R package bigleaf (version 0.6.5), an open source toolset for the derivation of meteorological, aerodynamic, and physiological ecosystem properties from eddy covariance (EC) flux observations and concurrent meteorological measurements. A ‘bigleaf’ framework, in which vegetation is represented as a single, uniform layer, is employed to infer bulk ecosystem characteristics top-down from the measured fluxes. Central to the package is the calculation of a bulk surface/canopy conductance (Gs/Gc) and a bulk aerodynamic conductance (Ga), with the latter including formulations for the turbulent and canopy boundary layer components. The derivation of physical land surface characteristics such as surface roughness parameters, wind profile, aerodynamic and radiometric surface temperature, surface vapor pressure deficit (VPD), potential evapotranspiration (ET), imposed and equilibrium ET, as well as vegetation-atmosphere decoupling coefficients, is described. The package further provides calculation routines for physiological ecosytem properties (stomatal slope parameters, stomatal sensitivity to VPD, bulk intercellular CO2 concentration, canopy photosynthetic capacity), energy balance characteristics (closure, biochemical energy), ancillary meteorological variables (psychrometric constant, saturation vapor pressure, air density, etc.), customary unit interconversions and data filtering. The target variables can be calculated with a different degree of complexity, depending on the amount of available sitespecific information. The utilities of the package are demonstrated for three single-level (above-canopy) eddy covariance sites representing a temperate grassland, a temperate needle-leaf forest, and a Mediterranean evergreen broadleaf forest. The routines are further tested for a two-level EC site (tree and grass layer) located in a Mediterranean oak savanna. The limitations and the ecophysiological interpretation of the derived ecosystem properties are discussed and practical guidelines are given. The package provides the basis for a consistent, physically sound, and reproducible characterization of biometeorological conditions and ecosystem physiology, and is applicable to EC sites across vegetation types and climatic conditions with minimal ancillary data requirements.}, + language = {en}, + number = {8}, + urldate = {2022-12-19}, + journal = {PLOS ONE}, + author = {Knauer, Jürgen and El-Madany, Tarek S. and Zaehle, Sönke and Migliavacca, Mirco}, + editor = {Bond-Lamberty, Ben}, + month = aug, + year = {2018}, + keywords = {Big-leaf model, land-atmosphere exchange}, + pages = {e0201114}, + file = {Knauer et al. - 2018 - Bigleaf—An R package for the calculation of physic.pdf:/Users/benjaminstocker/Zotero/storage/QANYDUCQ/Knauer et al. - 2018 - Bigleaf—An R package for the calculation of physic.pdf:application/pdf}, +} + diff --git a/data-raw/albedo_cescatti.csv b/data-raw/albedo_cescatti.csv new file mode 100644 index 0000000..a0a4a13 --- /dev/null +++ b/data-raw/albedo_cescatti.csv @@ -0,0 +1,54 @@ +site,country,vegtype,lat,lon,n_obs,albedo_modis,albedo_insitu +AU Tum,Australia,EBF,− 35.66,148.15,733,0.11,0.11 +AU Wac,Australia,EBF,− 37.43,145.19,275,0.09,0.1 +BR Cax,Brazil,EBF,− 1.72,− 51.46,67,0.12,0.12 +BR Sa3,Brazil,EBF,− 3.02,− 54.97,79,0.13,0.12 +BW Ghg,Botswana,SAV,− 21.51,21.74,28,0.16,0.18 +BW Ghm,Botswana,WSA,− 21.2,21.75,28,0.18,0.17 +BW Ma1,Botswana,WSA,− 19.92,23.56,252,0.16,0.14 +CA Ca1,Canada,ENF,49.87,− 125.33,580,0.09,0.09 +CA Ca3,Canada,ENF,49.53,− 124.9,552,0.13,0.14 +CA NS6,Canada,OSH,55.92,− 98.96,353,0.1,0.12 +CA SF2,Canada,ENF,54.25,− 105.88,363,0.11,0.11 +CA SF3,Canada,ENF,54.09,− 106.01,362,0.1,0.1 +CA WP1,Canada,MF,54.95,− 112.47,353,0.11,0.13 +CZ BK1,Czech Republic,ENF,49.5,18.54,148,0.09,0.1 +DE Geb,Germany,CRO,51.1,10.91,328,0.17,0.18 +DE Hai,Germany,DBF,51.08,10.45,451,0.13,0.13 +DE Kli,Germany,CRO,50.89,13.52,256,0.16,0.19 +DE Tha,Germany,ENF,50.96,13.57,477,0.1,0.07 +DE Wet,Germany,ENF,50.45,11.46,375,0.07,0.05 +ES ES2,Spain,CRO,39.28,− 0.32,298,0.14,0.13 +ES LMa,Spain,SAV,39.94,− 5.77,455,0.16,0.19 +FR Fon,France,DBF,48.48,2.78,162,0.14,0.13 +FR Hes,France,DBF,48.67,7.07,474,0.15,0.14 +FR Pue,France,EBF,43.74,3.6,401,0.11,0.12 +GF Guy,French Guyana,EBF,5.28,− 52.93,216,0.12,0.1 +HU Bug,Hungary,GRA,46.69,19.6,523,0.16,0.2 +IE Dri,Ireland,GRA,51.99,− 8.75,160,0.2,0.22 +IT Bon,Italy,ENF,39.48,16.54,174,0.1,0.08 +IT Col,Italy,DBF,41.85,13.59,183,0.14,0.15 +IT SRo,Italy,ENF,43.73,10.28,512,0.08,0.09 +JP Mas,Japan,CRO,36.05,140.03,177,0.12,0.13 +KR Kw1,Korea,MF,37.75,127.16,216,0.09,0.09 +NL Ca1,Netherlands,GRA,51.97,4.93,369,0.17,0.22 +NL Lan,Netherlands,CRO,51.95,4.9,108,0.17,0.18 +NL Loo,Netherlands,ENF,52.17,5.74,404,0.11,0.09 +PT Esp,Portugal,EBF,38.64,− 8.6,437,0.13,0.11 +RU Che,Russia,MF,68.61,161.34,104,0.15,0.16 +SE Nor,Sweden,ENF,60.09,17.48,234,0.1,0.09 +UK Gri,UK,ENF,56.61,− 3.8,15,0.1,0.09 +US Aud,USA,GRA,31.59,− 110.51,1244,0.18,0.21 +US Bn1,USA,ENF,63.92,− 145.38,104,0.11,0.09 +US Bo1,USA,CRO,40.01,− 88.29,721,0.16,0.19 +US Bo2,USA,CRO,40.01,− 88.29,289,0.16,0.19 +US Fmf,USA,ENF,35.14,− 111.73,247,0.1,0.12 +US FPe,USA,GRA,48.31,− 105.1,775,0.15,0.16 +US Fuf,USA,ENF,35.09,− 111.76,231,0.1,0.13 +US IB1,USA,CRO,41.86,− 88.22,349,0.15,0.16 +US Ivo,USA,WET,68.49,− 155.75,62,0.18,0.19 +US MMS,USA,DBF,39.32,− 86.41,639,0.13,0.13 +US MOz,USA,DBF,38.74,− 92.2,563,0.12,0.11 +US SRM,USA,WSA,31.82,− 110.87,826,0.17,0.16 +US WCr,USA,DBF,45.81,− 90.08,487,0.14,0.15 +ZA Kru,South Africa,SAV,− 25.02,31.5,447,0.15,0.15 \ No newline at end of file diff --git a/renv.lock b/renv.lock index 6b10abd..3211d0f 100644 --- a/renv.lock +++ b/renv.lock @@ -17,6 +17,14 @@ "Hash": "3e0051431dff9acfe66c23765e55c556", "Requirements": [] }, + "DEoptimR": { + "Package": "DEoptimR", + "Version": "1.1-3", + "Source": "Repository", + "Repository": "CRAN", + "Hash": "72f87e0092e39384aee16df8d67d7410", + "Requirements": [] + }, "KernSmooth": { "Package": "KernSmooth", "Version": "2.23-22", @@ -100,6 +108,17 @@ "Hash": "543776ae6848fde2f48ff3816d0628bc", "Requirements": [] }, + "bigleaf": { + "Package": "bigleaf", + "Version": "0.8.2", + "Source": "Repository", + "Repository": "CRAN", + "Hash": "8678f1d5d980de37c7bddc9de56e2079", + "Requirements": [ + "robustbase", + "solartime" + ] + }, "bit": { "Package": "bit", "Version": "4.0.5", @@ -1073,6 +1092,16 @@ "sp" ] }, + "robustbase": { + "Package": "robustbase", + "Version": "0.99-2", + "Source": "Repository", + "Repository": "CRAN", + "Hash": "bae2e53c94459ff147aef478eac6ee94", + "Requirements": [ + "DEoptimR" + ] + }, "rpmodel": { "Package": "rpmodel", "Version": "1.2.0", @@ -1195,6 +1224,16 @@ "geometries" ] }, + "solartime": { + "Package": "solartime", + "Version": "0.0.2", + "Source": "Repository", + "Repository": "CRAN", + "Hash": "cd75b243f7f10be541255f0801528722", + "Requirements": [ + "lubridate" + ] + }, "sp": { "Package": "sp", "Version": "2.1-2",