ALBEDO OPTIMIZATION

Select region: global amazon sahel siberia

STEP 1 : LEAF ALBEDO OPTIMIZATION

Data points selection (month-pixels from climatology 2009-2018) :

• MODIS is valid (not masked)
• ORCHIDEE is valid (not masked)
• Snow fraction = 0
• Total bare soil fraction < 0.5
• Continental fraction > 0.99
• Inland water fraction < 0.1

Optimization method: global L-BFGS-B with analytical gradient calculation (14 optimized parameters)

ndata - total number of used month-pixels, npts - total number of used grid pixels

LEAF_ALB_NIR	PFT2	PFT3	PFT4	PFT5	PFT6	PFT7	PFT8	PFT9	PFT10	PFT11	PFT12	PFT13	PFT14	PFT15
prior	0.23	0.18	0.18	0.2	0.24	0.15	0.26	0.2	0.24	0.27	0.28	0.26	0.24	0.24
min	0.17	0.17	0.1	0.1	0.16	0.1	0.1	0.16	0.1	0.17	0.1	0.15	0.1	0.2
max	0.25	0.24	0.23	0.24	0.24	0.19	0.33	0.23	0.36	0.36	0.46	0.36	0.29	0.35
post	0.204	0.189	0.15	0.176	0.228	0.129	0.263	0.178	0.296	0.258	0.259	0.292	0.222	0.323

LEAF_ALB_VIS	PFT2	PFT3	PFT4	PFT5	PFT6	PFT7	PFT8	PFT9	PFT10	PFT11	PFT12	PFT13	PFT14	PFT15
prior	0.04	0.04	0.04	0.04	0.03	0.03	0.03	0.03	0.06	0.06	0.06	0.06	0.06	0.06
min	0.02	0.02	0.01	0.01	0.02	0.01	0.02	0.02	0.02	0.04	0.02	0.05	0.02	0.04
max	0.06	0.08	0.09	0.08	0.07	0.05	0.04	0.04	0.18	0.18	0.23	0.15	0.16	0.1
post	0.0225	0.02	0.0231	0.0231	0.02	0.0234	0.02	0.02	0.0464	0.0485	0.0513	0.0606	0.02	0.04

---

STEP 2 : SNOW ALBEDO OPTIMIZATION

Data points selection (day-pixels from 2018) :

• MODIS is valid (not masked)
• ORCHIDEE is valid (not masked)
• Snow fraction > 0
• Continental fraction > 0.99
• Inland water fraction < 0.1

Optimization method: global L-BFGS-B with analytical gradient calculation (8 optimized parameters)

ndata - total number of used month-pixels, total npts - number of used grid pixels

SNOW_ALB_NIR PFT1 PFT4/5/7 PFT6/8/9 PFT10-15 snowa_aged_prior 0.5 0.3 0.3 0.5 snowa_aged_min 0.3 0.1 0.1 0.3 snowa_aged_max 0.7 0.5 0.5 0.7 snowa_aged_post 0.492 0.165 0.279 0.525 snowa_dec_prior 0.15 0.15 0.15 0.15 snowa_dec_min 0.05 0.01 0.01 0.05 snowa_dec_max 0.3 0.3 0.3 0.3 snowa_dec_post 0.122 0.0164 0.0435 0.054	SNOW_ALB_NIR PFT1 PFT4/5/7 PFT6/8/9 PFT10-15 snowa_aged_prior 0.5 0.3 0.3 0.5 snowa_aged_min 0.3 0.1 0.1 0.3 snowa_aged_max 0.7 0.5 0.5 0.7 snowa_aged_post 0.489 0.171 0.269 0.527 snowa_dec_prior 0.15 0.15 0.15 0.15 snowa_dec_min 0.05 0.01 0.01 0.05 snowa_dec_max 0.3 0.3 0.3 0.3 snowa_dec_post 0.12 0.021 0.038 0.0555 tcst_snowa_prior 10 tcst_snowa_post 13.02
SNOW_ALB_VIS PFT1 PFT4/5/7 PFT6/8/9 PFT10-15 snowa_aged_prior 0.5 0.3 0.3 0.5 snowa_aged_min 0.3 0.1 0.1 0.3 snowa_aged_max 0.7 0.5 0.5 0.7 snowa_aged_post 0.7 0.137 0.365 0.7 snowa_dec_prior 0.15 0.15 0.15 0.15 snowa_dec_min 0.05 0.01 0.01 0.05 snowa_dec_max 0.3 0.3 0.3 0.3 snowa_dec_post 0.229 0.01 0.01 0.125	SNOW_ALB_VIS PFT1 PFT4/5/7 PFT6/8/9 PFT10-15 snowa_aged_prior 0.5 0.3 0.3 0.5 snowa_aged_min 0.3 0.1 0.1 0.3 snowa_aged_max 0.7 0.5 0.5 0.7 snowa_aged_post 0.7 0.137 0.364 0.7 snowa_dec_prior 0.15 0.15 0.15 0.15 snowa_dec_min 0.05 0.01 0.01 0.05 snowa_dec_max 0.3 0.3 0.3 0.3 snowa_dec_post 0.218 0.01 0.01 0.119 tcst_snowa_prior 10 tcst_snowa_post 15

---

STEP 3 : BACKGROUND ALBEDO OPTIMIZATION

Data points selection (month-pixels from climatology 2009-2018) :

• MODIS is valid (not masked)
• ORCHIDEE is valid (not masked)
• Total_bare_soil * (1 - snow_fraction) > 0.5

Optimization method: pixel-level L-BFGS-B with analytical gradient calculation (1 optimized parameter at each pixel, varying albedo by ±0.1 from the prior)

Post-processing: filling non-optimized pixels (having small bare soil fraction and not covered with snow throughout the year) with the average value from 10 closest pixels (where contfrac > 0.99 and inlandwater < 0.1)

ndata - total number of used month-pixels (or day-pixels on the right graph), npts - total number of used grid pixels, zpts - total number of not used grid pixels

---

RESULTS (maps)

---

RESULTS (seasonal cycle across different regional masks)